Molecular and clinical determinants of survival following relapse

Molecular and clinical determinants of survival following relapse after curative treatment of stage IIIII colon cancer (CC). Results of the translational study on the PETACC 3 - EORTC 40993 - SAKK 6000 trial A. D. Roth, D. Klingbiel, P. Yan, R. Fiocca, M. Delorenzi, R. Labianca, D. Cunningham, E. Van Cutsem, F. Bosman, S. Tejpar Taiwan 2000 PETACC 2009 3 ASCO 2010

Rationale • Our previous results showed that stage II and III colon cancers harbour different biomarker alteration frequencies with different prognostic effects according to disease stage (ASCO proc. 2009, 27 abstr. #4002) • There is a lack of data regarding biomarkers prognostic for disease behaviour after relapse • Early observations suggest that marker sets prognostic for the risk of relapse might be different from those prognostic for survival after relapse (SAR) PETACC 3 ASCO 2010

BRAF status is prognostic in OS, very little or not in RFS! KRAS BRAF Relapse free survival (RFS) Overal survival (OS) Roth, A. D. et al. J Clin Oncol; 28: 466 -474 2010 Copyright © American society of Clinical Oncology PETACC 3 ASCO 2010

Objectives • To look for clinical and molecular markers prognostic for survival after relapse (SAR). • To assess their respective and relative prognostic relevance in the relapsing patient population using SAR as endpoint. • To assess possible differences between markers prognostic for risk of relapse and markers prognostic for SAR. PETACC 3 ASCO 2010

PETACC 3 trial: Patient Distribution Total Accrual: 3278 NO suitable material for biomarkers # relapsing 3241 Patients treated 1837 1404 598 392 (32. 5%) With suitable material for biomarkers (27. 9%) # relapsing PETACC 3 ASCO 2010

Methods (1) • FFPE tissue blocks prospectively collected and cut in 5 -20µ sections • Immunohistochemistry (IHC) – P 53: mouse m. Ab clone D 07, ABC Basic DAB Detection (Ventana medical systems) – SMAD 4: mouse m. Ab clone B 8 (Ig. G 1, Santa Cruz Biotechnology). Novocastra polymer detection kit – Thymidylate Synthetase (TS): Monoclonal antibody TS 106/4 H 4 B 1 (Ig. G 1, Zymed). DAKO En. Vision detection system – Telomerase (HTERT): Monoclonal antibody NCL-h. TERT (Ig. G 2, Novocastra). DAKO En. Vision detection system PETACC 3 ASCO 2010

Methods (2) • DNA was extracted with phenol/chloroform from normal and tumoral tissues after macrodissection of FFPE sections. • Molecular analysis: – Microsatellite Instability (MSI): assessed on 10 markers (BAT-25, BAT-26, D 5 S 346, D 2 S 123, D 17 S 250, BAT 40, TGF-ß RII, D 18 S 58, D 18 S 69, D 17 S 787) < 3 positive markers = MSI stable (MSS) ≥ 3 positive markers = MSI high (MSI-H) – 18 q. LOH: multiple SNPs typing by pyrosequencing on Normal and tumor DNA – KRAS exon 2 and BRAF exon 15: Allele specific real time PCR on Tumor DNA PETACC 3 ASCO 2010

Statistical methods • Prognostic value of the markers was analyzed by Cox regression for SAR and adjusted for stage • p values of the multivariate analysis taken from the full model (unless indicated otherwise) • Time to recurrence (TTR) of ≤ 18 months defined early relapse (ER), >18 months late relapse (LR) • Survival curves were computed according to the Kaplan-Meier method and compared using the logrank test PETACC 3 ASCO 2010

PETACC 3 trial: Relapser Distribution With material (biomarker set) Full set 392 990 Ø material 598 221 ER 171 LR 43. 6% 343 ER 255 LR 42. 6% PETACC 3 ASCO 2010

Relapsing Patients Characteristics Variables Full set (n=990) % (. 95 CI) Marker set (n=392) % (. 95 CI) Early relapse 564 57% (53. 860. 1) 221 56. 4% (51. 361. 3) Age <60 502 50. 7% (47. 553. 9) 185 47. 2% (42. 252. 3) Sex Female 432 43. 6% (40. 546. 8) 162 41. 3% (36. 446. 4) Stage III 860 86. 9% (84. 688. 9) 337 86% (82. 189. 3) Left side primary 636 64. 2% (61. 267. 2) 250 63. 8% (58. 868. 5) 526 53. 1% (5056. 3) 202 51. 5% (46. 556. 6) Treatment arm 5 -FU/FA PETACC 3 ASCO 2010

Univariate Analysis for SAR in the marker set (N=392) HR (95% CI) P-value TTR (ER/LR) 1. 55 (1. 21 -2. 00) 0. 0006 age 1. 14 (1. 00 -1. 29) 0. 045 sex 1. 13 (0. 88 -1. 44) 0. 33 Tumor Grade (G-34 / G-12) 1. 69 (1. 20 -2. 37) 0. 003 stage (III versus II) 1. 60 (1. 10 -2. 35) 0. 015 Tumor site (right/left) 1. 86 (1. 45 -2. 38) 7. 85 e-07 Treatment group 1. 09 (0. 86 -1. 38) 0. 49 MSI (MSI-H / MSS) 1. 05 (0. 65 -1. 70) 0. 84 Thymidilate synthetase 0. 89 (0. 64 -1. 22) 0. 46 SMAD 4 1. 27 (0. 98 -1. 65) 0. 07 p 53 0. 81 (0. 63 -1. 04) 0. 10 h. TERT 1. 33 (0. 95 -1. 86) 0. 10 18 q. LOH 0. 75 (0. 54 -1. 05) 0. 10 BRAF mut/wt 3. 63 (2. 41 -5. 47) 6. 15 e-10 KRAS mut/wt 1. 04 (0. 80 -1. 35) 0. 76 PETACC 3 ASCO 2010

Multivariate Analysis for SAR in the marker set (N=392) HR (95% CI) P-value TTR (ER/LR) 1. 60 (1. 23 -2. 09) 0. 0005 age 1. 00 (0. 99 -1. 01) 0. 98 sex 1. 24 (0. 95 -1. 62) 0. 11 Tumor Grade (G-34 / G-12) 1. 52 (1. 02 -2. 25) 0. 04 stage (III versus II) 1. 53 (1. 00 -2. 36) 0. 051 Tumor site (right/left) 1. 69 (1. 29 -2. 21) 0. 0002 Treatment group 1. 09 (0. 84 -1. 39) 0. 52 MSI (MSI-H / MSS) 0. 51 (0. 28 -0. 95) 0. 034 Thymidilate synthetase 0. 99 (0. 68 -1. 44) 0. 95 SMAD 4 1. 21 (0. 91 -1. 60) 0. 18 p 53 0. 96 (0. 73 -1. 26) 0. 76 h. TERT 1. 37 (0. 96 -1. 96) 0. 09 18 q. LOH 0. 86 (0. 60 -1. 24) 0. 43 BRAF mut/wt 3. 61 (2. 24 -5. 81) 1. 24 e-07 KRAS mut/wt 1. 13 (0. 85 -1. 51) 0. 40 PETACC 3 ASCO 2010

SAR according to BRAF mutation status Median survivals (95% CI): - BRAF mut: 7. 49 m (4. 8 -11. 2) - BRAF wt: 25. 2 m (21. 1 -29. 5) (p = 1. 9 e-11) PETACC 3 ASCO 2010

SAR according to primary tumor site Full set (n=990) Marker set (n=392) Median survivals (95% CI) LEFT side: 28. 4 m (26. 5 – 32. 3) 27. 6 m (22. 9 – 33. 6) RIGHT side: 16. 2 m (14. 4 – 18. 5) p=4. 52 e-14 16. 1 m ( 12. 6 – 19. 0) p=2. 71 e-06 PETACC 3 ASCO 2010

SAR after relapse according to TTR Full set (n=990) Marker set (n=392) Median survivals (95% CI) LR: 30. 5 m (27. 1 – 34. 7) 30. 0 m (24. 7 – 38. 4) ER: 18. 4 m (16. 4 – 20. 5) p=2. 18 e-08 17. 9 m ( 15. 5 – 20. 3) p=0. 0003 PETACC 3 ASCO 2010

Multivariate Analysis for SAR omitting BRAF in the marker set (N=392) HR (95% CI) P-value TTR (ER/LR) 1. 69 (1. 30 -2. 20) 9. 54 e-05 age 1. 09 (0. 95 -1. 25) 0. 21 sex 1. 20 (0. 92 -1. 57) 0. 17 Tumor Grade (G-34 / G-12) 1. 45 (0. 97 -2. 17) 0. 07 1. 51 (0. 98 -2. 32) 0. 06 Tumor site (right/left) 1. 76 (1. 34 -2. 30) 4. 39 e-05 Treatment group 1. 08 (0. 84 -1. 38) 0. 55 MSI (MSI-H / MSS) 0. 69 (0. 37 -1. 28) 0. 24 Thymidilate synthetase 0. 98 (0. 67 -1. 42) 0. 90 SMAD 4 1. 29 (0. 97 -1. 70) 0. 08 p 53 0. 99 (0. 76 -1. 30) 0. 94 h. TERT 1. 28 (0. 89 -1. 83) 0. 18 18 q. LOH 0. 81 (0. 57 -1. 17) 0. 27 KRAS mut/wt 0. 97 (0. 73 -1. 28) 0. 82 stage (III versus II) PETACC 3 ASCO 2010

SAR: MSI interactions with BRAF, Tumor Grade and Tumor Site Cox regression model type Variables HR (95% CI) P-value HR (95% CI), p-value Bivariate Multivariate 4 variables BRAF mut/wt 3. 95 (2. 65 -6. 09) MSI (MSI-H / MSS) 0. 71 (0. 43 -1. 18) 1. 91 (1. 33 -2. 75) 0. 0005 MSI (MSI-H / MSS) 0. 75 (0. 45 -1. 26) 0. 28 1. 87 (1. 45 -2. 41) MSI (MSI-H / MSS) 0. 75 (0. 46 -1. 23) Grade (G-34 / G-12) Bivariate 1. 02 (0. 63 -1. 65) Tumor site (right/left) Bivariate MSI (MSI-H / MSS) Grade (G-34 / G-12) Univariate Interaction 0. 94 1. 73 (1. 21 -2. 48) 0. 002 Tumor site (right/left) 1. 70 (1. 32 -2. 19) 4. 68 e-05 MSI (MSI-H / MSS) 0. 48 (0. 28 -0. 80) 0. 005 BRAFmut/wt 3. 75 (2. 44 -5. 75) NA 1. 38 e-09 5. 53 e-10 0. 46 (0. 15 – 1. 37) p=0. 16 0. 19 0. 49 (0. 17 – 1. 37) p=0. 17 1. 14 e-06 0. 50 (0. 17 – 1. 47) p=0. 21 0. 25 PETACC 3 ASCO 2010

Conclusions • Whereas BRAF status and tumor site were not prognostic for RFS of stage II-III colon cancer (Roth, A. D. et al. J Clin Oncol; 28: 466 -474 2010), BRAF status, tumor site and TTR are highly prognostic for colon cancer survival after relapse • Tumor stage at diagnosis and MSI status have a weaker impact on survival after relapse • BRAF status, tumor site and TTR should imperatively be used to stratify studies in metastatic colon cancer • The independent effects of tumor site and TTR might be indicative of additional as yet non-identified prognostically significant molecular markers PETACC 3 ASCO 2010

Thank You For All Your Efforts!! Austria, Belgium, Bulgaria, Croatia, Czech Republic, Denmark, Egypt, Finland, France, Germany, Greece & Cyprus, Hungary, Ireland, Iceland, Israel, Italy, Netherlands, Norway, Poland, Portugal, Russia, South Africa, Slovakia, Slovenia, Spain, Sweden, Switzerland, Taiwan, Turkey & UK We would also like to thank Pfizer for facilitating the execution and analysis of the PETACC 3 study PETACC 3 ASCO 2010