Animal models T-Cell Lymphomas: We are illuminating the darkest of tunnels. April 27-29, 2015 Bologna

Transcription

1 Animal models T-Cell Lymphomas: We are illuminating the darkest of tunnels April 27-29, 2015 Bologna GIORGIO INGHIRAMI Weill Cornell Medical College, New York, NY Center of Experimental Medicine and Research (CeRMS) University of Turin

2 Why we need reliable pre-clinical models? Only a handful number of compounds reaches the clinical arena (1 of 10,000). Small numbers of poorly characterized tumor cell lines that inadequately recapitulate human disease Xenografts do not predict for human effects, with 90% of novel antineoplastic drugs failing despite antitumor efficacy in classical preclinical models. Conventional GEM mouse models express/loose target genes in all cells under ectopic/forced promoters/enhancer, often with a constitutive modality Xenograft and/or GEM, PDX models do not fully recapitulate the human tumour environment Erroneous use and misinterpretation of preclinical data from cell lines and animal models. A poor appreciation of pharmacokinetics and pharmacodynamics, and the use of problematic endpoints and testing strategies. Preclinical testing rarely includes predictive biomarkers that, when advanced to clinical trials, will help to distinguish those patients who are likely to benefit from a drug. C. Glenn Begley & Lee M. Ellis: Nature 483, (29 March 2012)

3 What about reproducible PTCL models? Although several cell lines derived from T-ALL exist. Rare in vitro models for neoplastic post-thymic lymphocytes are available, with HTLV-I+, CTCL and ALK+ ALCL lines representing the exception. Few spontaneous (Roquinsan) or engineered (i.e. ITK-SYK, NPM- ALK) mouse models, faithfully reproducing their corresponding human counterparts, have been successfully used to define the pathogenetic mechanisms leading to T-cell transformation and/or design and validate therapeutic protocols.

4 Acquired Capabilities of Cancer Mast cell Lymphokines M1 M2 TCR BCR T B Chemokines Esosomes Douglas Hanahan and Robert A. Weinberg Cell. 2011, 44(5):646-74

5 Have we the right models?: 2D versus 3D in vitro

6 Rapidly evolving strategies for new mouse models

7 Humanized mouse models IL2 IL4 IL7 IL9 IL15 IL21 NK The NOD.Cg-Prkdc scid Il2rg tm1wjl /SzJ mice 2007 Shultz L.D. et al Nature Rev Immunol 2007, 7,

8 The PDTX models Humanized NSG mice MHC-I ko IL-6 IL-3, CSF2, Kit ligand Class II BAFF MHC-IA2 NSG mice Organoids

9 NOD/scid and NSG mice have been successfully used to study human CTL Meyer L.H. et al 2011 Clappier E. et al 2011 Krejsgaard T. et al 2010

10

11 Patient Derived Tumor Graft from fresh and/or cryopreserved tissue samples Tumor fragment implants in NSG mice 6 wks Serial tumorgrafts ALK+ ALCL T2 T3 T4 T5 T6 T7 8 wks Serial Tumorgrafts ALK- ALCL T2 T3 T4 T5

12 Source of cancer samples: Fresh versus viable frozen tissues? anno 2002 anno 2003 anno 2004 anno 2005 anno 2006 anno 2007 anno 2008 anno 2009 anno 2010 anno 2011

13 Can different implantation routes improve PDTX grafting? Sub cutaneous Intra venous Intra peritoneal Diagnostic sample Intra bone Intra liver Intra Spleen

14 Precision Therapeutic Medicine: new ideas at work Tumor expansion Biorepository HTP molecular stratification Molecular stratifiers - WES - RNAseq - RRBs - Proteomics

15 Representative ALCL tumorgraft expansion ALCL-1

16 ALCL time curves along serial tumorgraft passages ALCL-1 ALCL-3 ALCL-2 ALCL-4

17 ALCL tumorgraft on disseminate to secondary lymphoid and parenchymal organs A B C

18 Total body MRI scanning and therapeutic response assessments NSG 1st Treatment "3D - Doctor" Volume (mm 3 ) 2nd Treatment D.o.B. D.o.Inj Case # Drug Start End MRI Spleen Liver Kidney Skin Drug Start End MRI Spleen D.o.D. 10/07/201 72, ,42 484, , CTRL /08/ /06/ /08/2012 Marioni 2156 CTRL /08/ , ,24 342, , /12/201 10/08/ /06/2012 Marioni 2055 CEP 14gg 26/07/ /08/ /08/ , ,27 380, , /06/201 05/09/ /07/2012 Marioni 2101 CEP 14gg 08/08/ /08/ /08/ , ,94 629, ,62 CEP 7gg 27/08/ /09/ /09/ , /06/ /07/2012 Marioni 2103 CEP+JQ1 08/08/ /08/ /09/ , ,16 771, , /09/ ,29 05/09/201 2

19 Total body MRI scanning and therapeutic response assessments ALCL-PDT Day 0 ALCL-PDT day 14 ALCL-PDT day 14 treated

20 Do PDTX fully recapitulate their corresponding primary lesions

21 Primary ALCL and matched ALCL tumorgrafts display identical immuno-profiles A Primary ALCL-1 B ALCL-1-T3 CD30 CD30 ALK Perforin ALK Perforin C ALCL-1 ALCL-2 ALCL-3

22 SNP array identify analogous patient genomic defects in primary and corresponding tumorgrafts ALCL-1 ALCL-2 ALCL-3

23 Does clonal evolution take place in NSG models?

24 Zairis S et al:moduli spaces of phylogenetic trees describing tumor evolutionary patterns. In press Evolutionary modes in PΣ3. ALCL-3-PDT P T3_2 T3_3 T1_2 T5 T17 T11 T4 T2 T3 T8 T3_4 Shared L.B.F. > = 1 G E CD HC E F C G B D E B K NL B E ON M G OP O P M N L Prim ary Xenograft

25 Can we discover novel hits?

26 Gene expression profiling shows unique signature for each individual tumorgraft ALCL line M R B P C D D L

27 T2 T3 T4 T6 T8/164 T8/177 T9 250L T9 250S Gene expression signatures of ALCL-PDT identified unique subsets among PTCL patients T1 T2 T3 T4 T8 T9 T11 T1 T2 T3 T4 331 T4 334 T5 T6 Cumulative Survival ALCL-3 ALCL-1 ALCL-2 ALCL-2-PDT like ALK+ ALCL ALK- ALCL PTCL ALCL-3-PDT like Overall Survival

28 Host-lymphoma Interactions Human Tumor Content Primary RNA-Seq PDX RNA-Seq HUMAN Normal Host Content In Silico Graft/Host Separation Differential Expression Graft RNA-Seq (Human) Host RNA-Seq (Mouse) Human Tumor Content MOUSE Normal Host Content Human Host Signature Mouse/Human Host Comparison Enviroment Contribution

29 PDTX PDTX Human Mouse 300 Human Mouse CCL2 CCL2 CCL2 250 CCL2 CCL2 CCL CCR2 CCR2 CCR2 100 CCR2 CCR2 CCR H m H m VEGF ERK1/2 mtorc NFkB AKT VEGF Migration mccl2 mccr2 Erk1/2 NFkB2 M1 M2

30 NSG tumorgrafts respond to conventional treatments as matched primary ALCL 1th line CHOP ALCL -1 2nd line CHOP Anti-ALK(CEP28122) ALCL-1

31 The efficacy of SGN-35 is related on the tumor burden ALCL-1

32 YF STAT3 MEF Anti JAK1/2 inhibitors represent an alternative approach for pstat3+ ALK- ALCL 12 h 24 h 12 h 24 h Ruxolitinib JAK1 P-STAT3 STAT3 ACTIN PUH71

33 MRI Spleen Volume (mm 3 ) Variation from base line (% of volume) Treatment of ALK+ ALCL PDXT predicts clinical responses TRAF1 ALK d 21d + 14d treat Coiled Coil TRAF Domain Coil NPXY sequences Tyrosine Kinase Domain n n9 n45 n24 n24 n24 n11 n10 0

34 Conclusions: hurtles and problems Low grade lymphoma do not successful implant Time of engraftment can be exceedingly long Many lymphoma require host support The relationship with the immune system is hardly reproducible in human reconstituted mice PDTX do not propagate successfully in vitro

35 Conclusions: advantages and benefits Lymphoma derived PDT represent novel and powerful tools for investigating high-grade/end stage processes Humanized NSG mice should facilitate the creation of broad PDT libraries Lymphoma derived PDT need extensive genetic and functional analyses Molecularly annotated PDT are powerful models for testing new compounds/therapeutic strategies which can be provided to selected cohorts of cancer patients.

36 Acknowledgements WCMC Fabrizio Tabbo Maria Todaro Ramona Crescenzo Marcello Gaudiano Michela Boi Leandro Cerchietti CeRMS Roberto Piva Cristina Abele Rodolfo Marchiorlatti Katia Messana Indira Landra Nicoletta Chiesa Columbia University Raul Rabadan Francesco Abate Sakellarios Zairis Policlinico di Milano Antonino Neri Luca Agnelli Hugef, Torino Sivia Deaglio Tiziana Vaisitti IOSI Francesco Bertoni Michela Boi Ivo Kwee IRCC-Candiolo Enzo Medico Barbara Martinoglio The European T-cell Lymphoma Study Group Genetics-driven targeted management of lymphoid malignancies AIRC 5x1000