1. ROUND TABLE ON PRECLINICAL THERAPETUICS PANEL: Jon Kurie (M.D. Anderson) Eric Holland (MSKCC) Pier Paolo Pandolfi (MSKCC) Charles Sawyers (UCLA) Kevin Shannon (UCSF)

2. ROUNDTABLE FORMAT (1)Introduction (Shannon) (2)Summary of Discussions Related to Model Validation at Preclinical Therapeutics Meeting (Holland/Richon) (3)Overview of Models and Criteria for Preclinical Validation (Shannon, Kurie, Sawyers, Pandolfi) (4)Group Discussion

3. MOUSE NF1 LEUKEMIA MODEL NF1 encodes a GTPase activating protein for Ras Children with NF1 and heterozygous Nf1 mutant mice are predisposed to myeloproliferative disorders (MPD) and leukemias Murine and human leukemias show loss of normal Nf1/NF1 allele, and hyperactive Ras signaling Although homozygous Nf1 mutant mice die in utero from cardiac defects, Nf1-deficient fetal hematopoietic cells consistently induce MPD in irradiated hosts

4. FEATURES OF Nf1 MURINE LEUKEMIA MODEL human and myeloid disorders similar at genetic, biochemical, and cell biologic levels molecular pathogenesis well-characterized recipient mice are immunocompetent predictable subacute couse in recipients assays exist to measure biochemical effects of therapeutics on Ras signaling

5. P i GTP GDP SOS, GRF, GRP Neurofibromin p120 GAP Ras GTP Ras GDP Ras GDP Processing Growth Factor Ral GDS ? Raf ERK MEK PI3’-Kinase NF1 or RAS Mutation ? Cell Survival

6. PRECLINICAL/CLINICAL TRAILS OF TARGETED THERAPEUTICS Mouse: Farnesyltransferase inhibitor (Merck) MEK1 inhibitor (Pfizer) PI3 kinase inhibitor (Clapp lab with Lilly) anti-GM-CSF conjugate (with J. Perentesis UMN) Human: Recombinant erythorpoietin (Ortho/Amgen) CMA-676 (anti-CD33 conjugate) (Wyeth)

7. MODEL VALIDATION

8. Kras LUNG CANCER MODEL (Jacks) Latent mutant G12D allele of Kras activated somatically by homologous recombination Mice consistently develop multiple foci of lung adenocarcinoma by age 2-3 weeks, and die by age 8-10 months from pulmonary insufficiency Downstream effectors of RasGTP activated in lung lesions (e.g. ERK, P13K/Akt, MKK4) Thymic lumphomas also seen, but not colon or pancreatic cancers

9. PRECLINICAL TRIALS IN Kras MICE Aerosolized administration of chemotherapeutic and preventive agents by jut nebulizer (e.g. liposimal retinoic acid, recombinant adenoviral vectors, small molecules such MEK and PI3 kinase inhibitors) Systemic administration of Cox-2 inhibitor

10. HUMAN CLINICAL TRIALS Lung Cancer Chemoprevention Trials (1)retinoids in current and former smokers (2) EGFR inhibitor (SUGEN) (3)farnesyltransferase inhibitor (Schering)

11. PRECLINICAL VALIDATION creation of Cre/lox Kras allele will facilitate performing prevention trials (no paradigm for human lung cancer) hope to model human metastatic phenotype by generating Kras/p53 double mutants (Dr. Gigi Lazano) test agents known to be efficacious in human lung cancer (taxol, CDDP, navelbine) to validate mouse model

12. DISUCUSSION QUESTION How do you validate a mouse cancer model from the standpoint of preclinical therapeutics? What are the elements of a “validated” model?

13. DISUCUSSION QUESTION How have you been able to develop collaborations with pharmaceutical companies to test compounds in mice? What elements of a mouse model or preclinical project appeal to pharmaceutical companies and what are companies skeptical about?