The Therapeutic I mplications of EML4/ALK, ROS-1 and Other N ew Biomarkers

1. The Therapeutic Implications of EML4/ALK, ROS-1 and Other New Biomarkers Lyudmila Bazhenova, MD Associate Clinical Professor Lung Cancer Unit Leader UC San Diego Moores Cancer Center

2. Objectives • Review current state of targetable lung cancer biomarkers • Review and contrast clinical characteristics of patients with EML4-ALK, ROS 1, and KIFB5-RET fusion protein, testing strategies and agents with clinical activity.

3. Genomic Evolution of Lung Cancer

4. Mechanism of Action of ALK, ROS1 and RET Fusion Oncogenes • All three are receptor tyrosine kinases (RTK) • ALK and RET are capable of homodimerization and self (ligantindependend) activation • Mechanism of self activation of ROS1 is being debated • Downstream signaling via RAS/ERK (proliferation), and PI3K/AKT and JAK/STAT( resistance to apoptosis)

5. Testing for Fusion Oncogenes F IHC expression Break apart FISH Amount of protein on the surface of the cell RT-PCR

6. ALK Fusion Gene

7. ALK Fusion Gene Echinoderm microtubule associated protein-like 4 N C ? anaplastic lymphoma kinase Adapted from Soda et al. Nature; 2007.

8. ALK Fusion Variants Sasaki, European Journal of Cancer; 2010.

9. Methods of ALK Detection • FISH break apart • Pros: independent of FPE • Cons: if inversion involves a small locus of 2p it could be false negative; can not distinguish variants; cut of is 15% of nuclei with split signal; low throughput • RT-PCR • Pros: Rapid detection and identification of each unique variant • Cons: False negatives; Loss of RNA during de parafinization; has to be multiplexed, i.e probes to all known variants. Unknown variants will not be detected. • IHC • Pros: easy • Cons: several antibodies have been developed which look promising as a screening tool. No commercially available IHC in the US. • VENTANA just received an approval in China with 93% concordance with FISH, sensitivity 100%, specificity 98%

10. EML4-ALK Fusion • Patients:younger, non smokers, with adenocarcinoma, adenosquamous carcinoma and rarely SCC • Frequency:4% in all, 33% in EGFR negative never smokers • Biology: 16 EML4-ALK variants have been identified in NSCLC. Clinical significance of each variant is unknown. • Testing:Visys break apart FISH (>15% cells with split signal in 50 nuclei scored). • ALK PCR, IHC • Therapy: crizotinib 1Shaw AT, ASCO; 2010; 2Kris MG. ASCO 2011; abstract CRA7506. 3Rodig SJ, Clin CancerRes; 2009;15 Soda M, et al. Nature; 2007;448

11. Clinical Efficacy of Crizotinib Some degree of tumor shrinkage 90% ORR - 60.8% DCR - 82.5% at week 8 Median time to response 7.9 weeks (2.1 - 39.6 weeks) Median response duration 49.1 weeks Median PFS 9.7 months (95% CI 7∙7–12∙8) N=149 • Unknown… • How crizotinib compares to chemotherapy 1st line • QOL • OS • TTP Camidge, Lancet oncology 13, 2012

12. 1stLine or Second Line • No studies examining the best placement of the drug. • FDA approved the drug without mentioning the line of therapy. • One can make a leap of faith from EGFR inhibitors and use it in the first line. • Profile 1007 compared crizotinib to 2nd line chemotherapy • PFS 7.7 m vs. 3 m in favor of crizotinib (HR 0.49; 95% CI, 0.37 to 0.64; P<0.001) • RR 65% vs. 20 % in favor of crizotinib ( p<0.001) • OS not different, 64% of patients in chemotherapy arm received crizotinib • QOL: greater reduction of symptoms and delay in new symptoms on crizotinib arm. • Profile 1014 will compare crizotinib to 1st line chemotherapy. Shaw NEJM ;June 2013

13. PFS Shaw NEJM; June 2013.

14. Patient Related Outcomes Shaw NEJM; June 2013

15. Characteristics of Progression • Patients were allowed to stay on the study post progression if they continued to derive clinical benefi • Median duration of treatment 43.1 m (Range 0.1-136.8) • 69/149 patients had disease progression at the data cut off. • 39 continued to receive crizotinib for at least 2 weeks post progression • 12 of them did that for 6 months • Range of post progression treatment is 21 to 591 days. • Most common new sites of progression were brain ( N=10), lung (n=5), liver ( N=3) Camidge, Lancet oncology 13, 2012.

16. Duration of Initial Response and Post Progression Therapy

17. CrizotinibResistance • L1196M • L1152R • C1156Y • F1174L Sasaki Clinical Cancer Research. Epub2011.

18. Management of CrizotinibResistance • Local treatment with radiation for locally progressing disease • Clonal evolution • Platinum based doublet or triplet • Second generation ALK inhibitors • AP26114 • LDK378 • CH5424802 (RG7853) • HSP 90 inhibitors

19. Responses to Second Generation Inhibitors in crizotinibResistant Tumors • LDK378( phase I) 58% ORR1 • CNS penetration • CH5424802 ( phase I/II) • 48% ORR2 • CNS penetration • AP26113 ( phase I/II) • 76% ORR3 • CNS penetration 1Shaw. ASCO 2013 abstr 8010. 2Gadgeel, World Lung 2013, O16.06. 3Camidge. World Lung, MO0706

20. ROS1 Fusion

21. Oncogenic ROS1 • First described fusion gene FIG-ROS1 was found in glioblastoma • 240kb deletion on 6p21q resulting in a fusion gene coding for oncogenic fusion protein. • Short and long isoforms • Induce tumorigenesis in xenograft mouse models • Also expressed in cholangiocarcinoma in 8.7% and ovarian cancer in 0.5%, gastric and colon, myofibroblastic tumors and angiosarcoma • EZR–ROS1 fusion gene has been shown to promote lung adenocarcinoma when ectopically expressed in lung epithelium Gu TL, PLoS One; 2011. Birch AH, PLoS One; 2011; Lee, Cancer; May 2013 Bergethon et al. JCO; 2012 (30)8.

22. ROS 1 Fusion Gene Arai, PLOS ONE; February 2013.

23. ROS 1 Fusion Gene Variants

24. ROS1 Fusion • Patients:Younger, never smokers, adenocarcinoma, high grade histology • Frequency: 1.2 -1.7% in all • Biology: 9 variants have been identified in NSCLC so far • Clinical significance is unknown. Mechanism of activation is different. • FIG-, CD74-, SCL34A2-, TPM3-, SDC4-, EZR-, LRIG3, KDELR2–, and CCDC6– • Testing:Visys break apart FISH (>15% cells with split signal in 50 nuclei scored) • ROS PCR, IHC • Therapy: crizotinib Shaw AT, JCO 2012;30:(suppl; abstr 7508) Ou, Exprevi. of anticancer therapy 2012,;12 GuTL, PLoSOne. 2011; 6:e15640. Birch AH, PLoSOne. 2011; 6:e28250 Lee, Cancer May 2013 Davis ClinCancer Res . Sep 2012 Bergeron, JCO, 30, 2012

25. Methods of ROS1 Detection • RT-PCR • Cons: False negatives; 9 variants have been described in a matter of 12 months. Has to be multiplexed, i.e., probes to all known variants. Unknown variants will not be detected. • FISH break apart • Cons: if inversion involves a small locus it could be false negative; can not distinguish variants; cut of is 15% of nuclei with split signal; low throughput • IHC • Cons: not commercially available, several antibodies appear promising

26. Response of a ROS1 Positive Patient to Crizotinib • 49% homology in the TK domain and ATP binding site • Crizotinib is active in ROS1 fused cell cultures 12 weeks Baseline Bergeron, JCO, 30, 2012.

27. Clinical Validation of ROS1 as a Therapeutic Target • 14 patients enrolled in phase I study • Safety/efficacy of crizotinib 250mg bid • ROS1 rearrangement by FISH • Negative for ALK rearrangement • Average 54 yo, 13/14 never smokers • 80% received prior therapy • 8/14 responded (57%) Shaw et al. JCO. 2012, 30 (suppl; abstr7508.)

28. RET FUSION

29. Methods of RET Detection • RT-PCR • Cons: False negatives; 3 variants have been described in a matter of 12 months. Has to be multiplexed, i.e probes to all known variants. Unknown variants will not be detected. • FISH break apart • Cons: if inversion involves a small locus it could be false negative; can not distinguish variants; cut of is 15% of nuclei with split signal; not widely available; low throughput • IHC • Current IHC antibodies do not correlate with RET fusion

30. RET Fusion Gene

31. RET Fusion • Patients:AdenoCAand adenoSCC carcinoma, never or former smokers, poor differentiation ?, earlier LN metastases • Frequency: • 1.4% in all, • 5.6 % in “triple negative”( EGFR, ALK, KRAS) • 6.3% in non smokers negative for EGFR, KRAS, ALK, HER2, BRAF, and ROS1 • 16% in non smokers negative for EGFR, KRAS, ALK, ROS1, NRAS, BRAF, HER2, PIK3CA, MEK1, and AKT • Biology:4variants have been identified in NSCLC so far • Clinical significance is unknown. • KIF5B-, CCDC6-, NCOA4-. TRIM33 Ju YS, Genome Res, 2012 Drilon, CancerDiscoverMarch 2013 Wang R, J ClinOncol 30: 2012 Kohno, Cancer Science Aug 2013

32. RET Fusion Gene • Testing:Visys break apart FISH (> 15% cells with split signal in 50 nuclei scored) • RET PCR • Therapy: Unknown • Sunitinib, Sorafenib, Vandetanib, Carbozatinib, Ponatinib, and Lenvatinib all have potential for activity • All active in KIF5B-RET–transformed cell lines • Last 4 are in formal clinical trials

33. Clinical Activity of Carbozatinibin RET Fused Patients 4 weeks 4 weeks 4 weeks Drilon, Cancer Discover March 2013.

34. Summary

35. HER 2 Insertions • Patients:Adenocarcinomas, never smokers • Frequency:Incidence 2.8-4.2% • Biology: • In-frame insertions into exon 20. Transgenic mouse models confirm oncogenicity • Therapy: • Drugs of interest: neratinib, afatinib, dacomitinib • Preclinical models show synergy with mTORinhibitors. • Clinical trial of neratinib + temsirolimusongoing, several PR are reported • Both afatinib and dacomitinib have case reports of responses

36. BRAF Mutations • Patients: smokers and non smokers • Frequency: 1.6-3% • Biology: majority of the mutations are non V600E (more likely in smokers), V600E ( more likely in never smokers) • Therapy: • One case report or a NSCLC patient with V600E patient responding to vemurafenib • Dabrafenib is being tested in patients with V600E NSCLC • MEK inhibitors are being considered for non V600E patients

37. Q&A