microrna Signatures in Cancer E. ROBERT WASSMAN, MD, FAAP, FACMG CHIEF MEDICAL OFFICER, ROSETTA GENOMICS

microrna Signatures in Cancer E. ROBERT WASSMAN, MD, FAAP, FACMG CHIEF MEDICAL OFFICER, ROSETTA GENOMICS 1st International Congress on Controversies in Personalized Oncology (CONPO) BARCELONA, SPAIN MARCH 9, 2013 1

Safe Harbor Statement Except for historical information, the statements made in the following presentation, including without limitation, statements relating to projected cash burn rate and the expected timing of development of our pipeline products, and the future potential of microrna products in diagnostic and therapeutic markets are forward looking statements. Such forward looking statements involve significant risks and uncertainties that may cause actual results and events to differ materially and adversely from those implied by the forwardlooking statements. Risks and uncertainties include: that the timing and successful development of diagnostic and therapeutic products is highly uncertain; as well as the risks and uncertainties set forth under Risk Factors in Rosetta s Annual Report on Form 20 F for the year ended December 31, 2011, filed with the Securities and Exchange Commission. Rosetta is presenting this information as of the date of the presentation and expressly disclaims any duty to update the information contained in this presentation. This presentation contains information from third party sources, including data from studies conducted by others and market data and industry forecasts obtained from industry publications. Although Rosetta Genomics believes that such information is reliable, we have not independently verified any of this information and we do not guarantee the accuracy or completeness of this information. 2

micrornas: The Basics Discovered in 1993: central to cell differentiation & development in C.elegans micrornas are ~22nt long single stranded RNAs Post transcriptional regulation Master control at translation of protein coding genes Identified in humans in 2001 with ~2000 found since Linked to diseases, notably cancer, immune mediated disorders, viral infections, cardiovascular disesase, 4 neurological and metabolic disorders Winter et al., Nature Cell Biology 2009 3

Biologic role makes micrornas ideal cancer biomarkers & potential therapeutic candidates MicroRNAs are aberrantly expressed in all cancers studied MicroRNAs are oncogenes and tumor suppressors micrornas target self sufficiency in growth signals, apoptosis evasion, limitless proliferation, sustained angiogenesis and tissue invasion and metastasis MicroRNA genomic deletions correlate with specific cancers (Bcell lymphoma) Iorio and Croce, EMBO Molecular Medicine 2011 4

MicroRNAs are ideal biomarkers for cancer diagnostics Short RNAs embedded in protein complex: robust to degradation Stable markers in a variety of tissue samples and body fluids Stable at room temperature for years under FFPE conditions MicroRNAs identify tissues by source regardless of sample preparation Expression profiles classify most tumors with high accuracy and may predict clinical outcomes 2 years Equal quality 11 years correlation 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 2 3 3 5 7 7 8 8 8 9 9 10 11 age (years) 5 5

MicroRNAs identify tissues by source regardless of sample preparation microrna expression profiles cluster according to the patient/source whether sourced as fresh frozen or formalin fixed paraffin embedded (FFPE) tissues; However, mrna expression profiles most closely reflect the sample type 1 6 1. Liu A, et al. (2009) Int J Clin Exp Pathol. 2:519-527. 6 6

MicroRNAs demonstrate high tissue specificity Epithelial origin separateted from non epithelial origin using only 2 micrornas mir 200c and mir 205 Published by Rosetta in March 2008 1 Subsequent independent papers suggested same micrornas control epithelial mesenchymal transition (EMT) 1 Rosenfeld, N., R. Aharonov, et al. (2008). "MicroRNAs accurately identify cancer tissue origin." Nat Biotechnol 26(4): 462-9. 7

High tissue specificity translates into an accurate lung cancer assay using only small tumor samples Historically, in ~20% 30% of cases, specimen quantity & quality, and available methodology limits correct classification of lung tumors qrt PCR microrna analysis of 8 mirs clearly distinguishes between the 4 main subtypes of lung tumors Highly suitable for sub optimal Fine Needle Aspirates (FNA), brushings, and other small specimens 8

Lung cancer discrimination validated performance higher than histopathology Results of assay on 451 validation set samples Squamous Non Squamous Carcinoid Results of assay on 45 validation set samples Small Cell Failed Total Correctly classified 133 153 33 65 N/A 384 In correctly classified 6 13 2 5 N/A 26 Total 139 166 35 70 41 451 Overall Sensitivity: = 93.7% Overall Specificity: = 97.9% Sensitivity for pathology specimen: = 92.1% Sensitivity for cytology specimen: = 95% Overall success rate (samples with answer): = 91% Overall correct samples: 384/410 Reference: Classification of the four main types of Lung cancer using a microrna based diagnostic assay. Gilad.S, Lithwick Yanai G, Barshack I. et all. J Mol Diagn. 2012 Sep;14(5):510 7. 9

Similarly, high tissue specificity translates into an accurate renal cancer subtyping assay Fridman, E, et al: Accurate Molecular Classification of Renal Tumors Using MicroRNA Expression. J Mol Diag, 12(5):687 96, 2010. 10 10 10 10

Renal tumors assessed by KNN (5) algorithm using 24 micrornas Clear Cell Papillary Oncocytoma Chromophobe Clear Cell 53 1 Papillary 50 1 Chromophobe 39 3 Oncocytoma 1 4 32 Overall 174/184 classified correctly 95% accuracy 1 sample failed sample QA due to insufficient RNA; 16 samples: no results were generated (8%) 11

microrna based decision tree also sorts multiple cancer types effectively 12 12 Epithelial GI origin Non Epithelial origin SQ origin Adeno origin RCC Sarcoma Neuroendocrine Germ Cell

Microarray analysis of 64 mirnas can identify 42 distinct tumor origins Trained with ~1300 FFPE tumor samples Independent validation test set of 509 samples 146 (30%) were metastatic tumors Result of 1 or 2 tumor types generated from two algorithms Only 20 samples (4%) failed to produce result Overall, the correct tissue of origin was predicted by at least one of the reported answers in 418 (85%) ( sensitivity of prediction ) 403 samples (82%) received a single answer, of which 361 (90%) were classified correctly Overall specificity up to 99% 54 samples (11%) resulted in a unified categorical answer, e.g. Sarcoma, RCC, Thyroid type not specified, etc. (7 categories) Meiri E, Mueller WC, Rosenwald S, et al. A Second-Generation MicroRNA-Based Assay for Diagnosing Tumor Tissue Origin. Oncologist 2012 May 22 13 13

Most robust discrimination achieved via two discrete complementary classifiers Array results Signal Processing Decision Tree Two classifiers are employed to interpret the expression of 64 micrornas as a combined classifier KNN 0-5 Decision Maker -10-15 -15-10 -5 0 14 14

Studies in actual CUP patients at leading international centers show 80 92% concordance Prospective study on 75 CUP patients at M.D. Anderson done on first generation assay (narrower range of mirnas & tumor types) and results corroborated with second generation assay 84% concordance with final diagnosis 1 Two studies (first* & second** generation) on 52 real CUP patients as well as on metastases of known origin at Heidelberg University 80%* & 88%** concordance with final diagnosis 2,3 Study on 84 real CUP patients at University of Ioannina & Hellenic Cooperative Oncology Group 92% concordance with final diagnosis 4 1. Prospective Gene Signature Study Using microrna to Identify the Tissue of Origin in Patients with Carcinoma of Unknown Primary (CUP). Varadhachary, Spector et al. Clinical Cancer Research 17 2011 2. Meiri, E, Mueller, WC, Rosenwald, S, et al. A Second-Generation MicroRNA-Based Assay for Diagnosing Tumor Tissue Origin. The Oncologist, 2012; 17(6): 801-812. 3. Accurate classification of metastatic brain tumors using a novel microrna-based test. Muller, Spector et al. The Oncologist 16:165-74, 2011 4. G Pentheroudakis, N Pavlidis et al. A Novel microrna-based Assay demonstrates 92% accuracy in classification of metastatic tumors from patients diagnosed with carcinoma of unknown primary. Poster presentation at ASCO meeting in Chicago, June 2012 15

Ioannina University Hospital, Greece study of use in real CUP cases 1 Retrospective cohort of resected metastatic lesions from 84 CUP patient 1 Each prediction was assessed for agreement with: Pathological information Clinical information Follow up and outcome data 92% concordance overall Additional Dx and follow up Additional IHC for 4 patients Initial Dx process Presentation Clinical Dx Dx Clinical Dx Clinical Dx Final Clinical Dx Final Clinical Dx Tested Tested Patients 84 84 (100%) 59 (70%) Agreement 75 (89%) Agreement 77 (92%) Agreement mirview mets 2 Assay Results Assay Results 1. G Pentheroudakis, N Pavlidis et al. A Novel microrna-based Assay demonstrates 92% accuracy in classification of metastatic tumors from patients diagnosed with carcinoma of unknown primary. Poster presentation at ASCO meeting in Chicago, June 2012 16

Potential impact of microrna test results in Ioannina University Hospital CUP cases 1 In 77 patients (92%) test results were fully concordant with final diagnosis For 18 patients (21%), following test results would have resulted in administration of different chemotherapeutic regimens In 9 patients, different combination chemotherapies likely to be more active possibly with superior survival In 16 patients, the change in diagnosis could have been coupled to a change in targeted therapy employed In the remaining patients, results in agreement with the presenting diagnosis were still deemed useful objective information to narrow the differential diagnosis and increase the confidence level for treatment strategy 1. G Pentheroudakis, N Pavlidis et al. A Novel microrna-based Assay demonstrates 92% accuracy in classification of metastatic tumors from patients diagnosed with carcinoma of unknown primary. Poster presentation at ASCO meeting in Chicago, June 2012 17 17

Case report brain metastases in a woman with triple negative breast cancer (TNBC) A B H&E HMB45 Clinical presentation: Female presented with breast mass several months earlier Pathology: Dx TNBC Poorly differentiated adenocarcinoma with tumor giant cells, adenoid formations IHC: negative for Estrogen Receptor, Progesterone Receptor and Her2/neu S100 x100 x200 18 18

Case report triple negative breast cancer in actuality CUP from lost melanoma microrna microarray test result: Melanoma IHC after molecular profiling consistent with microrna: HMB45 positive S 100 focally positive MELAN A negative Breast lesion was then molecularly profiled on microrna array and resulted in the same diagnosis of melanoma IHC for breast lesion same as metastasis Published in: The Oncologist 16:165 74, 2011 Accurate classification of metastatic brain tumors using a novel microrna based test Muller, Spector et al. 19 19

The challenge of triple negative breast cancer (TNBC): can microrna open new paths? In TNBC cells, 13 micrornas recently reported as dysregulated versus normal 1 Metastatic signature of TNBC: 6 micrornas differentially expressed in tumor vs. metastasis and in normal vs. metastasis, but not in normal vs. tumor mir 424 is down regulated in mets vs. primary Aberrant mir 424 has been observed in progression of other cancer types, and may indicate a role in the development of metastases 2 1. Cascione L, Gasparini P, Lovat F et al. Integrated MicroRNA and mrna Signatures Associated with Survival in Triple Negative Breast Cancer. PLoS One. 2013;8(2):e55910 2. Xu J, Li Y, Wang F, Wang X et al. Suppressed mir-424 expression via upregulation of target gene Chk1 contributes to the progression of cervical cancer. Oncogene. 2013 Feb 21;32(8):976-87 20

microrna expression signatures may be able to stratify survival in TNBC 4 micrornas significantly associated with overall survival of TNBC patients Combined 4 mir signature stratified median overall survival of 69 vs. 83 month between the high and low microrna based risk groups Cascione L, Gasparini P, Lovat F et al. Integrated MicroRNA and mrna Signatures Associated with Survival in Triple Negative Breast Cancer. PLoS One. 2013;8(2):e55910 21

Ovarian Cancer Response To First Line Platinum Based Treatment micrornas from FFPE tumor samples of 57 ovarian cancer patients receiving platinum based therapy as first line hybridized to Rosetta s microarrays 7 micrornas were found to be significantly differentially expressed in tumors from platinum sensitive vs. platinum resistant patients 5 micrornas were associated with significant difference in survival or recurrence free survival High expression of hsa mir 27a identified a sub group with very poor prognosis survival hsa mir 27 expression correlates with survival Eitan R, Kushnir M, Lithwick-Yanai G, et al., Tumor microrna expression patterns associated with resistance to platinum based chemotherapy and survival in ovarian cancer patients. Gynecol Oncol. 2009 Aug;114(2):253-9 22

Mesothelioma diagnosis: orphan clinical unmet need of increasing importance 61 year old female presented with peritoneal carcinomatosis and ascites Presumptive diagnosis: primary peritoneal/ovarian carcinoma. Failed response to taxane platinum therapy, but had indolent course, with overall survival of 30 months Later microrna profile gave a single answer: mesothelioma Additional IHC tests confirmed diagnosis This guided diagnosis might suggest alternative therapy of pemetrexed/platinum Results of further IHC performed following microrna test results: A H&E B CK5/6 C Calretinin D Mesothelin G Pentheroudakis, N Pavlidis et al. A Novel microrna-based Assay demonstrates 92% accuracy in classification of metastatic tumors from patients diagnosed with carcinoma of unknown primary. Poster presentation at ASCO meeting in Chicago, June 2012 23 23

Mesothelioma prognosis: orphan clinical unmet need of increasing importance Despite indolent course in previous patient, mesothelioma remains highly lethal with Median TTP and survival after surgery 8 12m and 12 20m Currently, inability to forecast outcomes limits clinicians from directing aggressive therapy to the individuals who may actually benefit from such an approach Elevated mir 29c* Associated With Better Prognosis 24 24

Functional Consequences of mir 29c* Overexpression in Mesothelioma Cell Lines Mir29c* is higher in normal mesothelial cell lines compared with MPM Associated with a more differentiated phenotype Functional analyses data is compatible with the clinical findings Pass HI, Goparaju C, Ivanov S et al. hsa-mir-29c* is linked to the prognosis of malignant pleural mesothelioma. Cancer Res. 2010 Mar 1;70(5):1916-24 25 25

Elevated mir 29c* Associated With Better Mesothelioma Prognosis mir 29c* had significantly different expression (p = 0.00036, FC 1.8) between long and short TTP Elevated mir 29c* was associated with longer survival time in both training and test sets 11 10 9 8 7 6 Bad prognosis hsa-mir-29c* p-value 0.000355 fold change 1.8 Good prognosis Pass HI, Goparaju C, Ivanov S et al. hsa-mir-29c* is linked to the prognosis of malignant pleural mesothelioma. Cancer Res. 2010 Mar 1;70(5):1916-24 26 26

mirna biomarkers for stratification of response to novel immunotherapy Experimental immunotherapy trial in metastatic solid tumor 12 mirnas identified differentiating 10 responders from 10 nonresponders at Day30 (all passed FDR P<0.05) A classifier using two mirna s from day 30 markers also shows good signal at outset of therapy 24 PCR 23.5 hsa mir B hsa mir B 23 22.5 22 21.5 hsa mir A 21 22 22.5 23 23.5 24 24.5 hsa mir A 27

Looking Forward with micrornas Blood based diagnostics Additional Response/Predictive Biomarkers for New Drugs Identify patient sub cohorts that respond well to drugs stalled in development Response prediction suggested in several cancer types to date Therapeutics Up or down regulating a micrornas can regulate entire pathway Liver cancer (HCC): animal POC successfully completed & published Ovarian cancer: initial, pre clinical phase Recent studies in triple negative breast cancers suggest opportunity Non oncology applications 28