The Classification of Pancreatic Neuroendocrine Neoplasms: WHO 2017 - PowerPoint PPT Presentation

The Classification of Pancreatic Neuroendocrine Neoplasms: WHO 2017 Günter Klöppel, N. Volkan Adsay, Carlo Capella, Anne Couvelard, Ralph H. Hruban, David S. Klimstra, Paul Komminoth, Stefano La Rosa, Jean-Yves Scoazec, Nobuyuki Ohike, Robert Y. Osamura, Aurel Perren, and Guido Rindi

Definition: “Neoplasm” Overarching term to encompass all of the pancreatic entities with significant neuroendocrine differentiation (tumors, carcinomas, mixed carcinomas)

Classification of Pancreatic Neuroendocrine Neoplasms (WHO 2004)  Microadenoma (<0.5 cm)  Well differentiated endocrine tumor  Benign behavior: confined to pancreas, <2 cm, non-angioinvasive, </= 2 mitoses per 10 HPF, </= 2% Ki67-positive cells  Uncertain behavior: confined to pancreas >/= 2 cm, >2 mitoses per 10 HPF, > 2% Ki67-positive cells, OR angioinvasive  Well differentiated endocrine carcinoma  Low grade malignant: invasion of adjacent organs or metastases  Poorly differentiated endocrine carcinoma  High grade malignant: >10 mitoses per 10 HPF Kloppel et al. Ann NY Acad Sci 2004; 1014: 13-27

Classification of Pancreatic Neuroendocrine Neoplasms (WHO 2004): Issues  Combined staging (organ-confined, size) and grading (proliferative rate) parameters  Used both “tumor” and “carcinoma” to refer to the same entity  Changed diagnosis with disease progression  Used “carcinoma” for both well and poorly differentiated neoplasms  Provided no prognostic stratification for advanced disease  “Benign behavior” and “uncertain behavior” were NOT!

2006 ENETS Grading of GEP-NETs Grade Mitoses Ki-67 Index G1 <2 / 10 H.P.F. < 2% G2 2-20 / 10 H.P.F. 3-20% G3 >20 / 10 H.P.F. >20%  Poorly Differentiated (High Grade ) Neuroendocrine Carcinoma

Pancreatic NETs: Overall Survival by Grade Rindi et al., J Natl Cancer Inst 2012; 104: 764

TNM Staging System for Pancreatic Neuroendocrine Neoplasms (AJCC/UICC 2009) T – PRIMARY TUMOR TX Primary tumor cannot be assessed T0 No evidence of primary tumor T1 Tumor limited to the pancreas and size </= 2 cm T2 Tumor limited to the pancreas and size > 2 cm T3 Tumor extends beyond the pancreas but without involvement of the celiac axis or SMA T4 Tumors involves the celiac axis or the SMA (unresectable primary tumor)

Classification of Pancreatic Neuroendocrine Neoplasms (WHO 2010)  Well differentiated  Well differentiated neuroendocrine tumor, Grade 1 (NET G1)  Well differentiated neuroendocrine tumor, Grade 2 (NET G2)  Poorly differentiated  Poorly differentiated neuroendocrine carcinoma, Grade 3 (NEC G3)  TNM should be performed in all cases

WHO Grading of GEP-NETs (2010) Grade Mitoses Ki-67 Index G1 <2 / 10 H.P.F. < 2% G2 2-20 / 10 H.P.F. 3-20% G3 >20 / 10 H.P.F. >20%  High Grade (Poorly Differentiated) Neuroendocrine Carcinoma Virchows Archiv 2007, 451:757-762 Neuroendocrinology 2008, 87:1-64

WHO Grading of GEP-NETs: Provisions  Count mitoses in 50 high power fields  Assess Ki67 based on counting 2000 (500) cells  Assess Ki67 in “hot spots”  If mitotic rate and Ki67 are discordant, assign higher grade

What about G2 / G3 discordance?? (well differentiated tumor vs. poorly differentiated carcinoma )

Well Differentiated PanNET Mitotic rate = 8 / 10 HPF Mitotic rate = 12 / 10 HPF Ki67 = 45% Ki67 = 55%

Poorly Differentiated Neuroendocrine Carcinoma Chromogranin Ki67

Progression of Low Grade to High Grade Neuroendocrine Tumor Mitoses <1/10 HPF Mitoses 13/10 HPF

Ki67 = 2% Ki67 = 45% G1 G3 Tang et al., Clin Cancer Res 2016; 22: 1011

Poorly Differentiated Neuroendocrine Carcinoma of Pancreas

Genetics of Neuroendocrine Neoplasms of the Pancreas Gene Small Cell Large Cell W.D. Ductal ACa Small Cell NEC PanNET Lung CA KRAS 25% 33% 0% >90% 0-10% 11% 50% 0% 80-95% 0-10% CDKN2A TP53 100% 90% 4% 75% 80% 0% 10% 0% 55% 0% SMAD4 RB1 89% 50% 0% 13% 90% 0% 0% 43% 0% DAXX/ATRX MEN1 0% 0% 44% 0% 0% 15% 1% mTOR genes Yachida et al., Am J Surg Pathol 2012; 36: 173 Jiao et al., Science 2011; 331: 1199

Predictive and prognostic factors for treatment and survival in 305 patients with advanced gastrointestinal neuroendocrine carcinoma (WHO G3) • Reviewed clinical data on advanced stage G3 NECs, 2000-2009 • Ki67 > 20% • 252 patients received chemotherapy (platinum-based) • Median survival = 11 mos. • Response rate = 31% • Stable disease rate = 33%  Ki67 < 55% predicted a lower response rate (15% vs 42%, p < 0.001)  Ki67 < 55% predicted a better survival (14 vs 10 months, P < 0.001) Sorbye et al., Ann Oncol 2013; 24: 152-60

Conclusion: G3 NETs with Ki67 20-55% may be well differentiated biologically!! (“Well Differentiated PanNET with an Elevated Proliferative Rate”)

Survival of High Grade Neuroendocrine Neoplasms of the Pancreas Basturk et al., Am J SurgPathol 2015; 39: 683-690

Are all G3 Neuroendocrine Neoplasms the Same?  NO!  Small cell carcinoma vs. Large cell NE carcinoma  Large cell NE carcinoma vs. G3 well differentiated NET  NEC G3 vs. NET G3

Grading of Pancreatic Neuroendocrine Neoplasms: Proposal Well differentiated NE tumor* Poorly differentiated NE carcinoma* Grade Mitoses Ki-67 Index Grade Mitoses Ki-67 Index G1 <2 / 10 HPF <2% G3** >20 / 10 HPF >20% *Small cell carcinoma and large cell NE carcinoma; less organoid architecture, G2 2-20 / 10 HPF 3-20% classic cytology of small cell and large cell NE CA, absence of G1 or G2 NE components, may have non- G3** >20 / 10 HPF >20% neuroendocrine carcinoma components, *Organoid architecture, “well less diffuse immunoexpression of general differentiated” cytology, absence of non - NE markers neuroendocrine carcinoma components, **mitoses >20/10 HPF; Ki67 >20% and may have components of G1 or G2, usually >50% usually strong immunoexpression of general NE markers **mitoses usually <20/HPF; Ki 67 >20% but usually <50%

Classification of Pancreatic Neuroendocrine Neoplasms (WHO 2017)

Determining the Ki67 Labeling Index of NETs Courtesy of Dr. Laura H. Tang

Ki67 Cutpoints Grade Ki67 2010 Ki67 2017 G1 <2% <3% G2 3-20% 3-20% G3 >20% >20%

What about the G1/G2 cut-point??  Several studies have suggested 5% stratifies outcome better than 3% HOWEVER:  Statistical basis for defining cut-point is complex  Not all studies support the same cut-point  Currently no significant treatment change for G1 vs. G2  Changes in grading parameters confound historical data interpretation THERFORE:  Keep G1/G2 cut-point the same  Recommend reporting actual Ki67 index

G1 G2 G3 G3 WDNET PDNEC 0 10 20 30 40 50 60 70 80 90 100 Ki67%

How to distinguish G3 NEC (esp. large cell NE carcinoma) from G3 NET?

Pancreatic G3 NE Neoplasms Large Cell NEC G3 NET

How to distinguish G3 NEC (esp. large cell NE carcinoma) from G3 NET? • Clinical clues • History of well differentiated NET? • Octreotide scan positive? • FDG-PET positive? • Morphologic clues • Lower grade component? • Non-neuroendocrine component? • Mitotic rate? • Molecular clues • Status of TP53, RB1, DAXX, ATRX, MEN1

Well Differentiated PanNETs (G1-3) Exhibit a Different Molecular Phenotype from Poorly Differentiated NECs (G3) DAXX / TP53 Rb MEN1 ATRX WD- 4% 0 43% 44% PanNET PD- 56% 72% 0 0 PanNEC Jiao et al. Science 2011; 331: 1199 Yachida et al., Am J Surg Pathol 2012; 36: 173

PD-NEC PD-NEC WD-NET p53 Rb p53 Rb DAXX Tang et al., Am J Surg Pathol 2016; 40: 1192

Classification of 33 High Grade Pancreatic Neuroendocrine Neoplasms by Secondary Evidence Immunohistochemical Other Histologic Initial Consensus Abnormalities Components Confirmed Classification WD-NET G1/G2 WD-NET WD-NET WD-NET DAXX G1/G2 WD-NET WD-NET WD-NET ATRX G1/G2 WD-NET WD-NET WD-NET G1/G2 WD-NET WD-NET WD-NET DAXX G1/G2 WD-NET WD-NET WD-NET G1/G2 WD-NET WD-NET Ambiguous G1/G2 WD-NET WD-NET Ambiguous G1/G2 WD-NET WD-NET Ambiguous DAXX G1/G2 WD-NET WD-NET Ambiguous ATRX G1/G2 WD-NET WD-NET Ambiguous DAXX G1/G2 WD-NET WD-NET  19/33 (58%) of  18/19 (95%) Ambiguous G1/G2 WD-NET WD-NET Ambiguous ATRX WD-NET high grade (G3) morphologically Ambiguous DAXX G1/G2 WD-NET WD-NET Ambiguous DAXX G1/G2 WD-NET WD-NET pancreatic NE ambiguous high grade Ambiguous G1/G2 WD-NET WD-NET Ambiguous G1/G2 WD-NET WD-NET neoplasms were pancreatic Ambiguous G1/G2 WD-NET WD-NET Ambiguous G1/G2 WD-NET WD-NET morphologically NEneoplasms Ambiguous p53/Rb PD-NEC Ambiguous p53/SMAD4 Ductal adenocarcinoma PD-NEC ambiguous successfully classified Ambiguous p53/Rb PD-NEC Ambiguous p53/Rb PD-NEC Ambiguous p53 PD-NEC Ambiguous Undetermined PD-NEC-LCC DAXX G1/G2 WD-NET WD-NET PD-NEC-LCC Rb PD-NEC PD-NEC-LCC Ductal adenocarcinoma PD-NEC PD-NEC-SCC p53 Ductal adenocarcinoma PD-NEC PD-NEC-SCC Rb PD-NEC PD-NEC-SCC p53/Rb Ductal adenocarcinoma PD-NEC PD-NEC Rb PD-NEC PD-NEC p53 PD-NEC Tang et al., Am J Surg Pathol 2016; 40: 1192