3/31/2016
1
Immunohistochemistry on Limited Tissue Samples:
Do’s and Don’tsAndrew M Bellizzi, M.D.Department of Pathology
University of Iowa Hospitals and Clinicsandrew‐[email protected]
Disclosures
• I fretted over this . . .
• Last night I was sleepless . . . in Seattle
Disclosures
3/31/2016
2
Disclosures
• Otherwise, I have nothing to disclose
Goals
• To make an accurate, specific diagnosis with asfew immunostains as possible
• To provide (prognostic and) predictive info• To avoid less useful immunostains• To perform clinically valid immunohistochemistry• To triage tissue for other ancillary studies
Outline
• Competing interests: other ancillary studies• Technical aspects of immunocytochemistry• Enemies of the state: non‐specific immunos• Next‐generation immunohistochemistry• Favorite markers/panels
– Carcinoma of unknown origin– Lung adenocarcinoma vs. squamous cell carcinoma– Mesothelioma– Solid pancreatic tumors– Sarcoma– Lymphoma
3/31/2016
3
63‐year‐old woman with liver, lung, and adrenal masses; presumed lung primary based on FNA of liver lesion 4‐months prior; EGFR/ALK/ROS1wild‐type; progression on platinum‐based chemotherapy; biopsy for cancer mutation profiling
TTF‐1
3/31/2016
4
Taher,Thanks for raising this issue in the setting of our increasingly engaging in this era of personalized medicine.This was my case, as staff pathologist.I ordered the TTF‐1 because I believed it was necessary to the diagnosis. In fact, in this case, it was negative. I had already ordered it BEFORE your call to the hot seat. So, there was no communication breakdown.I was aware of the prior diagnosis. I read your note while I was evaluating the glass slides. I was aware that the primary driver of this biopsy was to perform NGS.There was actually much more tumor present in the immunostained slide than on the initial H&E, fortunately. I was pessimistic of the prospects for NGS success based on the initial H&E.I was acutely sensitive of the purpose of this biopsy, which is why I limited myself to ONE critical immunostain. Taking one 5 micron section off the tissue block (for IHC) DOES NOT compromise the ability to perform molecular, especially if the sections for molecular are taken simultaneously.
The anatomic pathologist is first and foremost responsible that a correct diagnosis is rendered. So, in this case I have to reasonably reassure myself that this is, in fact, metastatic lung adenocarcinoma.In this instance, both purposes should be served. Your note on the consult sheet WAS CRITICAL. If you hadn't made this comment, I would have performed a more extensive immunopanel.As always, communication is key.I would be happy to discuss this issue at greater length or in person.Best,Andrew
Tension Between Diagnosis and Predictive Testing is Increasing
• Use of IHC in pre‐op lung cytologic specimens
• >600% increase in IHC utilization
Ocque R, Tochigi N, Ohori P, Dacic S. Am J Clin Pathol 2011;136:81‐7.
Study Period Diagnosis of AdCAIHC Frequency
Diagnosis of SCCIHC Frequency
2000‐2004 14% (22/156) 11% (5/46)
2005‐2010 86% (134/156) 89% (41/46)
Ancillary Tests in FNA/Small BiopsyTumor Type Standard Applications Extended Applications
Breast Cancer ER, PR, HER2
Esophagus/Gastric AdCA HER2
Oropharygeal SCC and Head and Neck SCC of Occult Origin
p16 or HPV ISH
Lung Adenocarcinoma EGFR, ALKRET, MET, LKB1, PTEN
Colon Cancer KRAS, NRAS, BRAF, MSI PIK3CA, PTEN, EGFR CN
Mesothelioma
Thyroid Aspirate BRAF, KRAS
Pancreatic Cyst Aspirate Cyst fluid analysis KRAS, DNA content
Melanoma BRAF KIT
Sarcoma FISH or RT/PCR (for specific diagnosis)
Hematolymphoid Flow cytometry, IgH/TCR gene rearrangements, FISH (for diagnosis or prognosis)
p16 FISH
ROS1, PD‐L1NGS Multigene Panel
3/31/2016
5
Aisner DL, Sams SB. Diagn Cytopathol 2012;40:511‐24.
Gailey MP, et al. Cancer Cytopathol 2015. 123;30‐9.Knoepp SM, Roh MH. Cancer Cytopathol 2013. 121:120‐8.
Snow AM, et al. BMC Clin Pathol 2014. 14;30.
What’s Old is New
Hunt JL, et al. Diagn Cytopathol 1998;18:377‐80.
3/31/2016
6
Lessons Learned
• Do consider every biopsy of a tumor to be a potential molecular diagnostics specimen
• Do consider cutting unstained sections up‐front for molecular testing, if ordering IHC
• Do communicate with your clinical colleagues about priorities for the specimen
• Do not give up hope if you’ve exhausted a specimen: you may be able to recover DNA from routinely processed material
FNA of a Mural‐Based Gastric MassCell block
KIT on subsequent resectionDOG1
KIT KIT and DOG1 Intensity by FNA Method
• Cell blocks of 52 GIST, 24 LMS, 10 other
• No significant difference in number of cells in EUS vs CT cellblocks
• EUS‐FNA: FNA collected into CytoLyt (methanol‐based fixative)
• CT‐FNA: FNA collected into RPMI
• plasma and thrombin added to produce cell block, fixed in 10% formalin, processed for paraffin embedding
Hwang DG, et al. Am J Clin Pathol 2011;135:448‐53.
3/31/2016
7
The staining of cytologic preparations with a battery of antibodies to the various cell components or products is very costly and rarely rewarding . . . The results of ICC vary according to the batch of antibodies and the technical skills of the laboratories, the interpretationof the results is not always easy, and the problem with borderline positive stains is often perplexing.
Koss LG. The future of cytology. The Wachtel lecture for 1988. Acta Cytol 1990;34:1‐9.
We This Might Be So . . .
• Surgical Pathologists order IHC on FFPE tissue
• Cytopathologists may request IHC on:– Direct Smears
• Air‐dried– Unfixed– Post‐fixed
• Alcohol‐fixed– Unstained– Stained– Destained
– Cytospins– Monolayer Preparations– Cell Blocks
Cell Block
• Mandelbaum FS. Diagnosis of malignant tumors by paraffin sections of centrifuged exudates. J Lab ClinMed. 1917; 2:580.
Cell Blocks: Advantages and Disadvantages
• Advantages– Processed similarly to surgical pathology material– Facilitates multiple sections– Easy to store
• Disadvantages– Not amenable to on‐site adequacy assessment– May be pauci/acellular
• 10.6% of 246 lung/thoracic FNA’s (Rafael OC, et al 2014)• 57% of 76 consecutive EBUS FNA’s (Knoepp, Roh 2013)
3/31/2016
8
Performance in Other Cytology Specimen Types Potential Sources of Pre‐Analytic Variation
• Delay in fixation• Type, concentration, pH of fixative• Fixation time• Tissue Processing• Paraffin impregnation (paraffin melting point)• Block/slide storage
Engel KB, Moore HM. Arch Pathol Lab Med 2011;135:537‐43.
Potential Sources of Analytic Variation
• Antigen retrieval– Yes or no– Enzyme or heat‐induced
• For heat‐induced: buffer, pH, heat source
• Primary antibody dilution• Duration of primary antibody incubation• Detection chemistry (ABC, polymer‐based)
Goldsmith J. CAP Today Q&A. July 13, 2013.
Optimization and Validation
• Optimization: determination of provisional assay conditions, which is most often involves staining a single case or small number of cases at varying assay conditions
• Validation: testing and appropriate tissue set to determine analytic sensitivity and specificity, to reasonably assure that the test performs as expected
Goldsmith J. CAP Today Q&A. July 13, 2013.Fitzgibbons PL, et al. Arch Pathol Lab Med 2014;138:1432‐43.
3/31/2016
9
Goldstein NS, et al. Appl Immunohistochem Mol Morphol 2007;15:124‐33.
Recommendation
• Tissue should be fixed in 10% neutral‐pH, phosphate‐buffered formalin for a minimum of 8 hours. Non‐formalin‐based fixatives and or alternative fixation methodologies are strongly discouraged in regard to IHC, in large part because performance data are limited and extrapolation from formalin‐fixed data is unreliable.
Recommendation
• Antigen retrieval (AR) is presumed to “restore” the antigenicity after the formalin fixation. The parameters of an AR protocol must be balanced to match the unique length and type of tissue fixation of the individual laboratory and the characteristics of the individual antibody.
Antigen Retrieval
• Air‐dried smears theoretically should require (less or) no AR
• Alcohol‐fixed smears (coagulation, rather than cross‐linking, fixation) theoretically should require (less or) no AR
• Over‐AR produces high‐background staining and strong edge staining
• Under‐AR produces false negative IHC
3/31/2016
10
Fitzgibbons PL, et al. Arch Pathol Lab Med 2014;138:1432‐43.
Recommendations
• Laboratories must validate all IHC tests before placing them into clinical service
• For initial analytic validation of nonpredictivefactor assays, laboratories should test a minimum of 10 positive and 10 negative tissues
Recommendations
• “If IHC is regularly done on cytologic specimens that are not processed in the same manner as the tissues used for assay validation (eg, alcohol‐fixed cell blocks, air‐dried smears, formalin‐postfixed specimens), laboratories should test a sufficient number of such cases to ensure that assays consistently achieve expected results . . . The strength of evidence was inadequate to address the criteria and number of samples needed for validation with cytology specimens. If an assay has not been fully validated on cytologic specimens, laboratories may include a disclaimer in their report that results should be interpreted with caution.”
Controls
• “Unless the laboratory has a large bank of similarly prepared cytology material for positive and negative IC controls (non‐formalin‐fixed cytology specimen by the exact same preparation method as the current sample being tested), then criteria for having “proper controls” is not met and any interpretation of IC results should be suspect.”
Fowler LJ, Lachar WA. Arch Pathol Lab Med 2008;132:372‐83.
3/31/2016
11
Controls
• “In our experience, unstained direct smears in high quantities can be prepared using centrifuged cellular material for effusion specimens and these can be used for positive control ICC reactions.”
Knoepp SM, Roh MH. Cancer Cytopathol 2013. 121:120‐8.
Multiplexing
• Double‐staining, triple‐staining, etc.• Performing a second IHC on a previously IHC‐stained slide (if negative)
Dabbs DJ, Wang X. Diagn Cytopathol 1998;18:166‐9.
Lessons Learned
• Do perform ICC on cell blocks, if possible• Do examine cell block technique if “ICC performing suboptimally
• Do consider including cytology specimens in clinical IHC validations
• Do include similarly processed positive controls on ICC runs
• Do not perform ICC on other cytology specimen types unless the procedure has been specifically optimized for them (esp. antibody dilution, antigen retrieval)
3/31/2016
12
Myths
• CA125 is specific for Müllerian origin• CA19‐9 is specific for pancreatic origin• CK19 is specific for pancreatobiliary origin• MOC‐31 is an adenocarcinoma marker• RCC is specific for renal origin • Vimentin is specific for sarcoma
CA125 Expression Tumor Site (adenocarcinoma unless otherwise noted)
% Positive(at ≥10% cells staining with ≥ moderate intensity)
Ovary 98
Endometrium 90
Cervix 53
Pancreas 82
Gallbladder 63
Lung 56
Cholangiocarcinoma 51
Head and Neck (SCC) 40
Thyroid 39
Breast 27
Lung (SCC) 23
Melanoma, lymphoma, sarcoma, neuroendocrine, hepatocellular
0
Gremel G, et al. Histopathology 2014;64:293‐305.
CA19‐9 Expression
Tumor Site (n) (adenocarcinoma unless otherwise noted)
% Positive
Pancreas (26) 85
Breast, ductal (119) 48
Breast, lobular (10) 10
Lung (35) 69
Stomach (39) 56
Colon (25) 76
Ovary (29) 41
Kidney (45) 27
Mesothelioma <<1
Kaufmann O, et al. Histopathology 1996;29:233‐40.
3/31/2016
13
CK19 Expression Tumor Site (adenocarcinoma unless otherwise noted)
% Positive(at ≥10% cells staining with ≥ moderate intensity)
Cholangiocarcinoma 100
Pancreas 100
Gallbladder 100
Lung 100
Ovary 98
Breast 98
Colon 98
Stomach 98
Urothelial 95
Cervix 93
Lung (SCC) 91
Melanoma, lymphoma, hepatocellular 0
Gremel G, et al. Histopathology 2014;64:293‐305. MOC‐31
MOC‐31 Expression
Tumor Type (n) % PositiveSquamous cell carcinoma (110) 41
Mesothelioma (10) 0
Hepatocellular carcinoma (10) 20
Neuroendocrine carcinoma (5) 60
Lung adenocarcinoma (6) 100
Colon adenocarcinoma (6) 100
Urothelial carcinoma (11) 55
Neuroendocrine tumor, midgut (5) 100
Neuroendocrine tumor, pancreas (5) 80
Savage EC, Gailey MP, Bellizzi AM. Abstract at 2014 USCAP Annual Meeting.
Dx Algorithm: Poorly Differentiated Carcinoma in the Liver
HCC vs. ICC vs. MetastasisPerform Hep Par 1, MOC‐31(alternatively Arginase‐1, Claudin‐4)
Hep Par 1 (+)/ MOC‐31 (‐) HCC
Hep Par 1 (‐)/MOC‐31 (+) ICC or Metastasis
IHC menu for this application:
HCC Markers:‐ Hep Par 1‐ GPC3‐ pCEA/CD10 (canalicular)‐ Arginase‐1Non‐HCC Markers:‐MOC‐31‐ Claudin‐4
3/31/2016
14
73‐year‐old man with liver lesion, subtotally necrotic Hep Par 1 MOC‐31
Hepatocellular carcinoma
Large right retroperitoneal tumor
3/31/2016
15
13 cm tumor spanning the kidney and adrenal, expresses CD10, melan A, and inhibin
PAX8
PAX8 vs PAX2 ExpressionTumor Type PAX8 (% Positive) PAX2 (% Positive)
Clear cell 97 95
Papillary 100 76
Chromophobe 88 56
Collecting duct 71 43
Serous 98 55
Endometrioid 94 25
Clear cell 100 19
Transitional 67 11
Thyroid 91 0
Ozcan A, et al. Am J Surg Pathol 2011;35:1837‐47.Ozcan A, et al. Arch Pathol Lab Med 2012;136:1541‐51.
Laury AR, et al. Am J Surg Pathol 2011;35:816‐26.
3/31/2016
16
Vimentin Vimentin
Burkitt lymphomaMelanoma
Endometrioid Endometrial vs. Endocervical AdCA
Endometrioid Endometrial(n=30: ER, vim, CEA)(n=29: p16)
Endocervical(n=26: ER, vim, CEA)(n=23: p16)
ER 93% 38%Vimentin 97% 8%CEA 70%
(usually in squamous foci)96%
p16 97% (any staining)27% (≥50% cells stain)10% (100% cells stain)
96% (any staining)96% (100% cells stain)
McCluggage WG, et al. Int J Gynecol Pathol. 2002 Jan;21(1):11‐5. McCluggage WG, Jenkins D. Int J Gynecol Pathol. 2003 Jul;22(3):231‐5.
Vimentin p16
ER, PRUterine cervical mass
Results of immunopanel favor endometrial origin
Lessons Learned
• Do not extrapolate the results of differential‐specific markers beyond the differential (e.g., CK19, MOC‐31)
• Do not order CA125, CA19‐9, RCC (No No Never)• Vimentin??? If you must . . . sparingly . . . But please don’t tell me that you did . . .
3/31/2016
17
Next‐Generation Immunohistochemistry
• Mine developmental biology and molecular genetic literature to find:– Lineage‐restricted transcription factors– Markers identified by gene expression profiling – Protein correlates of molecular genetic events
• Bottom line: our markers keep getting better
Gene Expression Profiling Comparing Urothelial, Kidney, and Prostate Cancer
Higgins JPT, et al. Am J Surg Pathol. 2007 May;31(5):673‐80.
Breast cancer GCDFP‐15Mammaglobin
Historically, diagnostic armamentarium geared toward cytoplasmic or membranous differentiation markers; reduced sensitivity in poorly differentiated tumors
GATA‐3: highly expressed, regardless of differentiation
Primacy of lineage‐restricted transcription factors
3/31/2016
18
Regional Differential Diagnosis• Brain (carcinoma, melanoma, lymphoma, glioma)
• Lymph node (lymphoma vs. other)– Supraclavicular (anything), Periaortic (germ cell tumor)
• Mediastinum (lung, lung, lung, thymic [KIT,CD5], germ cell)
• Visceral organ (1° vs metastasis) – Lung (AdCA vs SCC vs metastasis) – Liver (HCC vs pancreatobiliary vs other)
• Pleural effusion (AdCA vs mesothelioma)
• Peritoneum/Ovary (1° surface epithelial vs metastatic GI)
• Retroperitoneum (well and dedifferentiated liposarcoma)
• Somatic soft tissue (sarcoma vs. metastasis)
• Bone (blastic‐prostate, breast; lytic‐RCC, thyroid; mixed‐lung)
Diagnosis of Broad Tumor Class
• Carcinoma (broad spectrum keratins, EMA [aka MUC1])
• Melanoma (S‐100, SOX10, melan‐A, HMB‐45, MiTF, BRAF)
• Lymphoma (CD45; CD43, CD79a, MUM1, ALK1, CD30 if LCA‐)
• Sarcoma (CD34, MDM2/CDK4 – DDLPS , add’l based on morph.)
• Mesothelioma (WT‐1, calretinin, CK5/6, D2‐40)
• Germ cell tumor (SALL4, PLAP)• Non‐epithelial neuroendocrine neoplasm (CG, SYN)
Diagnosis of Carcinoma Type
Carcinoma (cohesive, keratin and/or EMA+)
Adenocarcinoma(gland forming/
mucin producing) – see next algorithm
Squamo‐transitional(p63, p40, CK5/6+)
SCCCK7 var.GATA3‐ ? UC
CK7/CK20+GATA3+
Large polygonal cell(CK7/CK20‐)
HCCHep Par 1 +GPC3+
RCCPAX8+
AdCCMel‐A+Inhibin+SYN+SF1+
Neuroendocrine(CG and/or SYN+)
NETKi‐67 ≤20%
NECKi‐67>20%
Which Screening Keratin Should I Use?Clone 1 2 3 4 5 6 7 8 9 1
011
12
13
14
15
16
17
18
19
AE1/AE3 X X X X X X X X X X X X
OSCAR X X X X
MAK‐6 X X X X X X
MNF116 X X X X
CAM5.2 X X
KL1 X X X X X X X X X X X
34βE12 X X X X
• In general, any of these are acceptable• It’s not the number of keratins, per se, but rather the affinity
(e.g., CAM5.2 vs AE1/AE3 in HCC/RCC) • Stratified epithelia: K1‐6, 9‐17• Simple epithelia: 7, 8, 18, 19, 20
Ordóñez NG. Hum Pathol. 2013 Jul;44(7):1195‐215.
3/31/2016
19
Coordinate Expression of CK7/CK20Site (tumor) CK7 CK20Prostate, HCC, AdCC, RCC - -Lung, Breast, Müllerian,Pancreatobiliary (PB), Upper GI (UGI) + -Bladder, UGI, PB, Mucinous Ovarian,Colon (esp. Rectum), Occ. Lung + +Colon, Merkel cell, Occ. UGI - +
Cancer Epidemiology and Morphology‐Based Site of Origin Generator
MostCommon Primary Sites in Men (Rank Order)
Morphology MostCommon Primary Sites in Women(Rank Order)
Morphology
Prostate (25%) (AdCA)
Characteristic nuclear features regardless of Gleason grade: monomorphous and prominent nucleoli
Breast(26%)(AdCA)
Variable
Lung (15%) (AdCA, SCC, NEC)
Variable (AdCA) Lung (14%)(AdCA, SCC, NEC)
Variable (AdCA)
Colorectum (10%) (AdCA)
Characteristiccytoarchitectural features: “tall, dark, and dirty”
Colorectum(10%)(AdCA)
Characteristiccytoarchitectural features: “tall, dark, and dirty”
IHC to Assign AdCA Site of Origin: Ranked from Most to Least Ordered (2011)
• CDX2 (enteric differentiation)• TTF‐1 (lung, thyroid)• ER and PR (breast, Müllerian)• PAX8 (kidney, Müllerian, thyroid)• p53 and WT‐1 (serous carcinoma, latter also exp. by mesothelioma)• Napsin A (lung, papillary RCC)• PSA and PSAP (prostate)• GCDFP‐15 and mammaglobin (breast)• Thyroglobulin (thyroid)‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐‐• GATA‐3 (breast; also urothelial) • NKX3.1 (prostate)
Transcription factors in bold
Lung Adenocarcinoma vs Squamous Cell CarcinomaMarker AdCA
(% positive)SCC (% positive)
CK7 100 68TTF‐1 86 21p63 31 100CK5/6 21 100
Ocque R, Tochigi N, Ohori P, Dacic S. Am J Clin Pathol 2011;136:81‐7.
• p63, CK5/6 have a specificity problem• TTF‐1 has a sensitivity problem
3/31/2016
20
p40 (ΔNp63)
Bishop JA, et al. Mod Pathol 2012;25:405‐15.
Marker AdCA(% positive)
SCC (% positive)
Large Cell Lymphoma (% positive)
p63 31 100 54p40 3* 100 0* Each 1‐5% cells staining
• TTF‐1+/p40‐ AdCA• TTF‐1‐/p40‐ AdCA• TTF‐1‐/p40+ SCC
Napsin A
Turner BM, et al. Arch Pathol Lab Med 2012;136:163‐71.
Marker Lung AdCA(% positive)
SCC (% positive)
Napsin A 87 (n=303) 3 (n=200)TTF‐1 64 (n=94) 2 (n=94)
• Lower rate of TTF‐1+ AdCA than literature (75‐85%)• Napsin A frequently expressed by RCC (esp.
papillary) and clear cell carcinoma
Soft Tissue: IHC to Assign LineageLine of Differentiation Characteristic Historically Useful Immunostains
Adipocytic Fatty S‐100
Fibroblastic/myofibroblastic Spindled; light pink SMA, CD34, ALK
So‐called fibrohistiocytic Shrinking group (e.g., TSGCT) None especially
Smooth muscle Spindled; brightly eosinophilic SMA, desmin
Pericytic Spindled; “circumferential perivascular growth pattern”
SMA
Skeletal muscle Rhabdomyo(sarco)ma Desmin, myogenin
Vascular Vascular channels CD34, CD31, Factor VIII
Chondro‐osseous Cartilage, osteoid S‐100
Interstitial cell of Cajal GIST KIT, CD34
Nerve sheath Spindle cell; wavy nuclei S‐100, GFAP
“Uncertain” “Distinctive” tumor types (e.g., synovial sarcoma, epithelioidsarcoma, PEComa)
Keratins, EMA, S‐100, (TFE3, HMB‐45, WT‐1)
Undifferentiated/unclassified Formerly MFH Diagnosis of exclusion
Soft Tissue Markers Validated at U of Iowa in Last 3‐Years
Name Diagnostic ApplicationBeta‐catenin Desmoid fibromatosis (70%), solid‐pseudopapillary neoplasm,
pancreatoblastoma; hepatoblastoma (50‐90%), fetal‐type lung AdCA, colonic adenoma/AdCA
MDM2 and CDK4(MDM2 FISH)
Well and dedifferentiated liposarcoma
TLE1 Synovial sarcoma; Pitfalls: weaker staining in other sarcomas, frequently positive in carcinomas
MUC4 Low‐grade fibromyxoid sarcoma, sclerosing epithelioidfibrosarcoma (75%); also broadly expressed by epithelia
ERG Vascular tumors, prostate cancer (50%); occ. myeloid leukemias and Ewing sarcoma (10%)
STAT6 Solitary fibrous tumor (NAB2‐STAT6 translocation)
HHV8 Kaposi sarcoma, primary effusion lymphoma, multicentric Castleman disease, plasmablastic lymphoma arising in plasmablastic lymphoma
3/31/2016
21
Soft Tissue Markers Validated at U of Iowa in Last 3‐Years
Name Diagnostic Application
SOX10 Melanoma, MPNST (50%), clear cell sarcoma, myoepithelial differentiation; superior in sensitivity and specificity to S‐100
INI1 Epithelioid sarcoma (90%), malignant rhabdoid tumors of soft tissue, kidney, CNS (>95%), medullary carcinoma of the kidney, other INI1‐deficient tumors (some epithelioid MPNST, myoepithelial CA of soft tissue, extraskeletal myxoid chondrosarc.)
c‐Myc c‐Myc
DLBCLBurkitt lymphoma
Advantages of ICC over Flow Cytometry
• Hodgkin lymphoma• Burkitt lymphoma vs DLBCL, GC Type
– BL: c‐Myc+, Bcl‐2‐– DLBCL: c‐Myc‐, Bcl‐2+ or ‐
Tumor TypeCD20 CD79a PAX5
CD10 Bcl‐6
CD5 CD43Cyclin D1
Bcl‐2Other Useful Markers/Notes
Tumors composed of small cells Extranodal marginal zone lymphoma of mucosa‐associated lymphoid tissues (MALT lymphoma)
+ ‐ ‐ Occ. (30%) ‐ Var.
Kappa/lambda light chain restriction (occ.); keratins to highlight lymphoepithelial lesions
Mantle cell lymphoma + ‐ + + + +
Ki‐67 proliferation index >40‐60% associated with poor prognosis; rarely CD5 or cyclin D1‐
Follicular lymphoma + + ‐ ‐ ‐ +Higher grade tumors more likely to show aberrant IHC (e.g., CD10‐ or CD43+)
Chronic lymphocytic leukemia/small lymphocytic lymphoma
+ ‐ + + ‐ + CD23+
3/31/2016
22
Tumor TypeCD20 CD79a PAX5
CD10 Bcl‐6
CD5 CD43Cyclin D1
Bcl‐2Other Useful Markers/Notes
Tumors composed of intermediate/large cells
Diffuse large B‐cell lymphoma
+
CD10 Var. (30‐60%)
Bcl‐6+ (60‐90%)
Rare (10%) Occ. (25%) ‐ Var. MUM1 (35‐65%)
Burkitt lymphoma + + ‐ + ‐ ‐Ki‐67 proliferation index approaches 100%; TdT‐; c‐Myc+; Bcl‐2 occ. weak +
Plasmablastic lymphoma
CD20/PAX5‐
CD79a+ (50‐85%)
+ ‐ + ‐ +EBV EBER (60‐75%); CD45‐; MUM1/CD138/CD38+; EMA/CD30 var.; association with HIV
Granulocytic Sarcoma
‐ ‐ ‐ + ‐ ‐ CD68/KIT/CD99/ CD34/TdT Var.
B‐lymphoblastic lymphoma
CD20 Var. (25‐50%)
CD79a/PAX5+
CD10+ (60%)
Bcl‐6‐‐ + (67%) ‐ + TdT (95%), CD34/CD99+;
CD13/CD33 Occ.
T‐lymphoblastic lymphoma
CD20/ PAX5‐
CD79a Rare (5‐10%)
CD10 Var.
Bcl‐6‐Var. + (90%) ‐ +
TdT (95%), CD3/CD99+; CD4/CD8 double+ (70%); CD1a (67%); CD34 Var.; CD13/CD33 Occ.
Hematolymphoid Markers Validated at U of Iowa in Last 2‐Years
Name Diagnostic Application
Tryptase Mast cells
CD163 Monocyte‐macrophage lineage marker (more specific than CD68)
c‐MYC Burkitt lymphoma and c‐MYC activated DLBCL
SOX11 Mantle cell lymphoma (w/ emphasis on identifying CycD1‐ cases)
Lessons Learned• Do perform a limited panel of IHC to assign tumor type/site of origin based on clinical and morphologic differential (“big 3” + appropriate “next‐gen” markers)
• Do not equate p63+ with SCC; do consider p40
• Do consider using “next‐gen” markers to make specific sarcoma and lymphoma diagnoses
Thank you!!!
3/31/2016
23
U of Iowa Development QueueName Diagnostic ApplicationAndrogen receptor Sebaceous CA, salivary duct CA, prostate CA; Pitfall: low specificity
Arginase Hepatocellular CA; Pitfall: occ. stains AdCA
Claudin‐4 Broad‐spectrum epithelial marker not expressed by hepatocytes/meso’s
Clusterin Follicular dendritic cell sarcoma, ALCL, tenosynovial GCT, non‐ileal NET
D2‐40 Mesothelioma (inc. sarc.), lymphatic endothelium, seminoma, others
Glutamine synthetase Focal nodular hyperplasia, BCAT‐HA, HCC vs. B9 liver: Pitfall: other CAs +
Mammaglobin Breast CA; surrogate for ETV6 FISH in mammary analogue secretory CA
NKI‐C3 (CD63) Cellular neurothekeoma; + in many other tumors inc. melanoma
NUT NUT midline carcinoma (undifferentiated CA with abrupt keratinization)
p40 (ΔNp63) Squamous cell carcinoma (superior spec. to p63); basal cells and myoep’s
SATB2 Colon AdCA; osteoblastic differentiation; rectal NET
SDHB SDHB‐deficient GIST, paraganglioma/pheochromocytoma, other
SMAD4 Pancreatic ductal AdCA (50% demonstrate loss); midgut NET (40%)
SF1 Adrenal cortical neoplasms, sex‐cord stromal tumors
Diagnosis of Broad Tumor Class
Morphologic “Boxes”
• Cohesive, poorly cohesive, dyscohesive
• Spindle cell (sarcoma, sarcomatoid carcinoma)• Pleomorphic (anything)• Round cell (lymphoma, sarcoma)• Epithelioid (carcinoma, melanoma)
• Monomorphous or pleomorphic
Pleomorphic and High‐Grade are NOT Synonymous
3/31/2016
24
Monomorphous Pleomorphic
Clear cell sarcoma Melanoma
CDX2 homogenous: LGICDX2 heterogenous: UGI, mucinous ovarian, PB (20%)
CK20 diffuse, strong: LGICK20 weak, patchy: UGI, mucinous ovarian, PB (≤40%)
NET Site of Origin Algorithm
Maxwell JE, Sherman SK, Stashek KM, O’Dorisio TM, Bellizzi AM, Howe JR. Surgery. 2014 Dec;156(6):1359‐66.
3/31/2016
25
Midgut NETs have a characteristic histologic appearance and nearly always express CDX2
Pancreatic NETs are of diverse histologic appearance and nearly always express ISL1
CK20+/TTF‐1‐(≥90% Merkel cell carcinoma
TTF‐1+/CK20‐(≥90% SCLC, 45% visceral NEC)
Small cell lung carcinomaMerkel cell carcinoma
Transcription Factor “Infidelity” in NECFrequently Expressed Transcription Factors (%)
Merkel CellCarcinoma (n=40)
Small CellLung Carcinoma (n=24)
ExtrapulmonaryVisceral NEC (n=19)
FLI1 90 88 78GATA3 24 8 40Islet 1 97 88 58Myc 29 25 61PAX6 78 29 26PLAG1 58 79 72SATB2 63 42 89SOX2 100 96 89TTF1 5 96 44
Czeczok TW, Gailey MP, Hornick JL, Bellizzi AM. Mod Pathol. 2014 Feb;27(S2):152A.
NECs expressed a median of 8 TFs (range 0‐18) out of 38 examined
3/31/2016
26
76‐year‐old man with past tobacco use and asbestos exposure p/w pleural effusion; at thoracotomy diaphragm encased by nodular fibrous tissue; multiple plaque like areas throughout pleural cavity Pan‐keratin
Calretinin, WT‐1, CK5/6, MOC‐31, Ber‐EP4, STAT6
Mesothelioma‐Marker Expression in Sarcomatoid Mesothelioma
AE1/AE3 75% (18/24)CAM5.2 96% (23/24)MNF‐116 100% (21/21)Calretinin 25% (6/24)WT‐1 33% (8/24)D2‐40 100% (24/24)
Chirieac LR, et al. Am J Cancer Res. 2011;1(1):14‐24. All of these were prospectively dx’d as pancreatic neuroendocrine tumor . . . only one of them is
3/31/2016
27
Cellular Epithelioid Neoplasms of the Pancreas
• Solid neoplasms composed predominantly of neoplastic elements with little stroma– Pancreatic neuroendocrine tumor (PNET)– Solid‐pseudopapillary neoplasm (SPN)– Acinar cell carcinoma (ACC)– Pancreatoblastoma (PB)
• On cytology: dyshesive, monomorphic– SPN: 3 of 6 misdiagnosed as PNET (Bardales et al ‘04)– ACC: 2 of 4 misdiagnosed as PNET (Stelow et al ‘06)– ACC: 14/29 misdiagnosed, 5 as PNET, 5 as ductal AdCA
(Sigel et al ‘13)
Immunophenotype of Cellular Epithelioid Neoplasms of the Pancreas
PNET SPN ACC PB
Broad Spectrum Keratins
+ 70% + +
Synaptophysin 95% 20‐70% Rare cells 80%
Chromogranin 90% ‐ Rare cells 80%
Trypsin Rare cells ‐ 95% 95%β‐catenin <5% >95% 10% >90%
BCAT
Trypsin BCAT Trypsin
Core biopsy from a 40 cm retroperitoneal tumor demonstrates undifferentiated neoplasm composed of sheets of epithelioid cells.After performing 17 immunostains a diagnosis of “malignant neoplasm” indeterminate for sarcoma, carcinoma, or lymphoma was rendered.
3/31/2016
28
MDM2, CDK4
Most consistent with dedifferentiated liposarcoma Resection of similar case
Abrupt transition
Well‐differentiated component
Dedifferentiated component
Keratin and/or EMA‐positivity do not assure a diagnosis of carcinoma
Potential Pitfall #1
3/31/2016
29
Leiomyosarcoma Smooth muscle actin, desmin+ Keratin AE1/AE3+: keratin, EMA expression in 30‐40% LMS
33‐year‐old woman with 1‐year h/o R hip pain; large SQ mass
EMA+, AE1/AE3+, GATA‐3 focal + p63, CK5/6, ER, PR, TTF‐1, WT‐1, CD31, S‐100, HMB45 all ‐Conclusion: Favor metastatic carcinoma, ? breast or urothelial
3/31/2016
30
INI1: absent expression epithelioid sarcoma, proximal‐type63‐year‐old man with increasing hip pain x 1 month; proximal femur lesion with soft tissue extension
Undifferentiated Malignant Neoplasm with Osteoclast‐like Giant Cells
• Undifferentiated/anaplastic carcinoma– Keratin AE1/AE3– CDX2, PAX8, TTF‐1
• Osteosarcoma– SATB2
• Leiomyosarcoma– Desmin, SMA, caldesmon
Keratin AE1/AE3+ (desmin, SMA ‐)Conclusion: Favor undifferentiated carcinoma
3/31/2016
31
Musculoskeletal radiologist sug. presence of chondroid matrix (MRI)Subsequent femoral head resection for pathologic fracture
Abrupt transition from WD chondrosarcoma to undifferentiated neoplasm dedifferentiated chondrosarcoma
Keratin‐Positive Soft Tissue Tumors• Chondroid lipoma• Pleomorphic liposarcoma• Desmoplastic fibroblastoma• Solitary fibrous tumor• Inflammatory myofibroblastic tumor• Myxoinflammatory fibroblastic sarcoma• Leiomyosarcoma• Rhabdomyosarcoma• Schwannoma (cross‐reactivity with GFAP)• Epithelioid hemangioma• Pseudomyogenic hemangioendothelioma• Epithelioid hemangioendothelioma• Angiosarcoma• Gastrointestinal stromal tumor• Sclerosing perineurioma• Dermal nerve sheath myxoma
• Epithelioid MPNST• Ectopic hamartomatous thymoma• Ossifying fibromyxoid tumor• Myoepithelial tumors of soft tissue• Synovial sarcoma• Epithelioid sarcoma• Desmoplastic small round cell tumor• Extrarenal rhabdoid tumor• Undifferentiated/unclassified sarcoma• Chondroblastoma• Dedifferentiated chondrosarcoma• Conventional osteosarcoma• Ewing sarcoma• Chordoma• Adamantinoma• Osteofibrous dysplasia
EMA‐Positive Soft Tissue Tumors• Pleomorphic liposarcoma• Calcifying aponeurotic fibroma• Lipofibromatosis• Dermatofibrosarcoma protuberans• Solitary fibrous tumor• Low‐grade fibromyxoid sarcoma• Sclerosing epithelioid fibrosarcoma• Leiomyosarcoma• Pleomorphic rhabdomyosarcoma• Epithelioid hemangioma• Epithelioid hemangioendothelioma• Angiosarcoma• Neurofibroma• Perineurioma• Dermal nerve sheath myxoma
• Solitary circumscribed neuroma• Meningioma• Hybrid nerve sheath tumor• Acral fibromyxoma• Angiomatoid fibrous histiocytoma• Myoepithelial tumors of soft tissue• Synovial sarcoma• Epithelioid sarcoma• Desmoplastic small round cell tumor• Extrarenal rhabdoid tumor• Undifferentiated/unclassified sarcoma• Conventional osteosarcoma• Chordoma• Epithelioid hemangioma• Adamantinoma
3/31/2016
32
78‐year‐old man p/w word‐finding difficulty: L temporal lobe mass Pan‐keratin S‐100
Dr. Bellizzi, what kind of S‐100‐positive carcinoma is this?
S‐100 Expression in AdenocarcinomaPrimary Tumors Metastatic Tumors
Salivary gland 80% (n=15) 75% (n=4)
Lung 7% (n=27) 12% (n=25)
Breast 60% (n=20) 62% (n=8)
Esophagus 0% (n=8) 0% (n=2)
Stomach 20% (n=10) 25% (n=8)
Gallbladder 0% (n=1) 0% (n=1)
Colorectum 25% (n=28) 23% (n=13)
Pancreas 0% (n=8) 0% (n=5)
Kidney 65% (n=23) 66% (n=3)
Endometrium 78% (n=36) 64% (n=14)
Ovary 84% (n=24) 87% (n=22)
Prostate 0% (n=27) 0% (n=8)
Unknown origin 22% (n=9)
Total 43% (n=228) 39% (n=122)
Herrera GA, et al. Am J Clin Pathol. 1988 Feb;89(2):168‐76. MiTF
Melan A (A103)
Conclusion: metastatic melanoma
BRAFmutation testing: wild‐type
HMB‐45
3/31/2016
33
Keratin‐Positivity in MelanomaPrimary cutaneous melanoma (n=62)
Recurrent or metastatic melanoma (n=22)
Vimentin 100% 100%S‐100 95% 95%NSE 87% 77%HMB45 97% 64%NKI‐C3 97% 95%Cytokeratin(KL1, CAM5.2, 35βH11)
0% 23%
Achilles E, Schröder S. Pathologe. 1994 Aug;15(4):235‐41.
Keratin‐Positivity in Lymphoma
• Anaplastic large cell lymphoma (5/18; 28%)1
• 13/866 (1.5%) hematolymphoid tumors in TMA2
– 5/18 (28%) mantle cell lymphomas
• Otherwise, case reports and small series
1. Gustmann C, et al. Am J Pathol. 1991 Jun;138(6):1413‐22.2. Adams H, et al. Pathol Res Pract. 2008;204(8):569‐73.
EMA‐positivity in Hematolymphoid Neoplasms
• Anaplastic large cell lymphoma (50‐95%)• Plasma cell neoplasms (most)• Diffuse large B‐cell lymphoma
– TC/HR, ALK+, plasmablastic, primary effusion lymphoma
• T‐cell lymphoma (20%)• Nodular lymphocyte‐predominant Hodgkin (L&H)• Follicular dendritic cell sarcoma (var.)
Beware of LCA‐negative hematolymphoid neoplasms
Potential Pitfall #2
3/31/2016
34
LCA‐Negative Hematolymphoid Neoplasms
• Acute megakaryoblastic leukemia• B‐cell acute lymphoblastic leukemia/lymphoma• ALK‐positive large B‐cell lymphoma
– EMA+, CK‐/+; CD20‐, CD79a‐• Plasmablastic lymphoma (EBER usually +)
– EMA+/‐; CD20‐, PAX5‐• Anaplastic large cell lymphoma, ALK‐positive (some)• Classical Hodgkin lymphoma• Follicular dendritic cell sarcoma