Early tumor shrinkage ETS and depth of response

Early tumor shrinkage (ETS) and depth of response (Dp. R) to anti-EGFR in (m)CRC helpful surrogates or meaningless endpoints? Marc Peeters MD, Ph. D Coordinator Multidisciplinary Oncological Center Antwerpen (MOCA) Head of the Oncology Department UZA, Professor in Oncology UA

o Demographics - female patient, 56 years - well-controlled arterial hypertension o Current History - invitation for flemish colorectal screening program - i. FOB test : positive o Oncological History - total colonoscopy : semicircular lesion at the splenic flexure - pathology : moderately differentiated adenocarcinoma - staging : § CEA – 5. 6 µg/l (3. 5 µg/l) § thoracic/abdominal CT – no distant metastases non-metastatic adenocarcinoma of the left colon, p. G 2 T 3 N 0(0/16)c. M 0

o Follow-up - colonoscopy after 1 year : no lesions - every 3 months – clinical evaluation and CEA-level determination - every 6 months – liver ultrasound/thoracic radiology 2013 Date 11/01 24/04 09/07 23/10 CEA (µg/L) 2, 3 2, 7 2, 5 3, 2 2014 07/01 21/02 5, 7 11, 2

Hemoglobin Alk. Phosph. AST ALT LDH CEA (g/dl) (U/l) (µg/l) 12. 4 173 34 43 1367 14507 (12. 0 -15. 0) (53 -141) (< 34) (84 -246) (< 3. 0) - colonoscopy: stenotic tumor at 25 cm (sigmoid) - biopsy: moderately differentiated (G 2) adenocarcinoma

§ m. CRC – (clinical) heterogenuous population Resectable Neo-Adjuvant Potential Resectable Induction Limited Metastatic Control Diffuse Palliative Response

§ EGFR inhibitors in m. CRC – combination therapy Treatment PFS, mo OS, mo Chemo + bevacizumab 9. 4 -10. 6 20. 3 -21. 3 Chemo + anti-EGFR 7. 9 -11. 4 17. 0 -28. 4 FIRE-3 Chemo + bev vs Chemo + anti-EGFR 10. 2 10. 4 25. 6 33. 1 CALGB 80405 Chemo + bev vs Chemo + anti-EGFR 11. 3 11. 4 31. 2 32. 0 Adapted from Cremolini et al. Nat Rev Clin Oncol. 2015; doi: 10. 1038/nrclinonc. 2015. 129

Douillard JY et al. N Engl J Med 2013; 369: 1023 -1034

§ Assessment of Efficacy in First line m. CRC - PFS • Influenced by toxicity of the regimen and size of the tumour lesion, favouring larger lesions - OS • Influenced by 2 nd and subsequent lines of treatment • With longer post-progression survival, the influence of further treatment lines becomes greater - ORR: • Influenced by RECIST measurements • Used as an endpoint in Phase 2 studies Slide courtesy of S. Stintzing

Tumour change, % § Response – continuous measure based on RECIST SD range: ‒ 29 % to +19 % 20 SD - 30 ORR Slide courtesy of V. Heinemann SD = stable disease

Surrogate endpoint = - a biomarker intended to substitute for a clinical endpoint (NIH, 2001) - a laboratory measurement or physical sign used as a substitute for a clinical meaningful endpoint that measures directly how a patient feels, functions, or survives and that is expected to predict the effect of therapy (FDA, 1992) Earlier Measurement Smaller Sample Size Zhoa F. J Clin Oncol 2016; 34: 1436 -1437. Editorial

§ ETS – % decrease in tumour load at a given time Baseline = 100% ETS = % decrease (8 weeks) 8 weeks With a 20% cut-off, this results in a categorical decision: Early tumour shrinkage (ETS) Non-early tumour shrinkage (non-ETS) Heinemann V et al. Eur J Cancer 2015; 51: 1927 -36 Time

§ Additional (response) assessments – parameters (1) OS Lethal tumour load TTG Tumour load at baseline Dp. R ETS Tumour nadir PFS + molecular markers e. g. CEA, CA 19 -9 Time since start of treatment Heinemann V et al. Eur J Cancer 2015; 51: 1927 -36 Dp. R, depth of response; TTG, time to tumour growth

§ Additional (response) assessments – clinical (2) ETS: § Rapidly predicts sensitivity to treatment 1− 5 § Potential to relieve tumour-related symptoms 5 § May increase possibility of resection/chance of cure in some patients with initially unresectable metastases 5 ETS, Dp. R: § Potentially prognostic for PFS/OS 4− 6 1. Giessen C, et al. Cancer Sci 2013; 104: 718− 24; 2. Piessevaux H, et al. Ann Oncol 2009; 20: 1375– 82; 3. Piessevaux H, et al. J Clin Oncol 2013; 31: 3764− 75; 4. Stintzing S, et al. Ann Oncol 2014; 25(Suppl 5): v 1–v 41: abstract LBA 11 (and oral presentation); 5. Douillard JY, et al. Eur J Cancer 2015 ; 51: 1231− 42; 6. Petrelli F, et al. Eur J Cancer 2015; 51: 800− 7

FOLFIRI+Cmab Piessevaux H et al. J Clin Oncol 2013; 31: 3764 -3775 FOLFIRI FOLFOX+Cmab FOLFOX

Piessevaux H et al. J Clin Oncol 2013; 31: 3764 -3775

Douillard JY et al. Eur J Cancer 2015; 51: 1231 -42

§ PRIME – ETS for PFS and OS Panitumumab + FOLFOX 4 ETS at Week 8, RAS wild type < 20% ≥ 20% n (%)† 61 (28) 158 (72) 96 (43) 125 (57) 6. 7 (5. 4– 9. 9) 13. 6 (12. 0– 15. 7) 6. 1 (5. 3– 8. 0) 9. 9 (8. 0– 11. 1) Median PFS, months (95% CI) HR (95% CI) P-value 0. 62 (0. 45– 0. 85) 0. 0031 Phi coefficient‡ Median OS, months (95% CI) HR (95% CI) P-value Phi coefficient§ Douillard JY et al. Eur J Cancer 2015; 51: 1231 -42 0. 67 (0. 50– 0. 88) 0. 0040 0. 31 12. 6 (9. 3– 18. 2) 32. 5 (28. 3– 37. 6) 15. 2 (11. 4– 17. 2) 0. 47 (0. 34– 0. 65) < 0. 0001 26. 0 (22. 1– 31. 3) 0. 50 (0. 37– 0. 66) < 0. 0001 0. 34

§ PRIME – change in EQ-5 D health scores (baseline to discontinuation) WT RAS: symptomatic† Mixed effect model Difference P-value 0. 025 0. 070 0. 02 (0. 049, 0. 142) (‒ 0. 020, 0. 071) (0. 013, 0. 127) (n = 82) (n = 109) 4. 164 0. 300 4. 314 (1. 405, 7. 824) (‒ 2. 822, 3. 422) (0. 636, 7. 993) ETS ≥ 30% ETS < 30% HSI LS means (n = 84) (n = 109) 0. 096 (95% CI) OHR LS means (95% CI) 0. 02 †Patients with tumour-related symptoms at baselinedefined as EQ-5 D pain/discomfort scale score > 1. . EQ-5 D, Euro. Qo. L 5 domain; HSI, health state index; LS, least squares; OHR, overall health rating. § In patients with tumour symptoms at baseline, there were statistically significant improvements in Qo. L in those with ETS vs those without § These data add to the idea that achieving early reductions in tumour load is associated with symptomatic benefits for patients Siena S, et al. ESMO Open 2016; 1: e 000041

§ PRIME – ETS, conclusions - Whole population vs. Individuals - ETS in symptomatic patients & in conversion Douillard JY et al. Eur J Cancer 2015; 51: 1231 -42

§ FIRE-3 – median tumour diameter over time Median tumour diameter (95% CI) of patients not PD, % WT RAS 100 75 50 Bevacizumab + FOLFIRI Cetuximab + FOLFIRI 0 Cetuximab + FOLFIRI Bevacizumab + FOLFIRI 6 12 103 95 109 126 Stintzing S, et al. Ann Oncol 2014; 25(Suppl 5): v 1–v 41: abstract LBA 11 (and oral presentation) Week 22 32 79 99 45 43

§ FIRE-3 – additional response assessments Cetuximab + FOLFIRI (n = 199) Bevacizumab + FOLFIRI (n = 201) Median PFS, months 10. 3 10. 2 Median OS, months 33. 1 25. 0 ORR, † % (n = 157) 72. 0 (n = 173) 56. 1 (1. 27− 3. 19) P = 0. 003 Assessment 1 (n = 157) (n = 173) P-value ETS ≥ 20% at Week 6, % 68. 2 49. 1 ETS < 20% at Week 6, % 31. 8 50. 9 - Median Dp. R, % 48. 9 32. 3 P < 0. 0001 (n = 230) 83. 0 63. 0 (n = 242) 72. 3 47. 9 P = 0. 003 P = 0. 005 Assessment 1 CEA reduction 2# Maximum − median, % ≥ 75%, % HR/OR (95% CI) P-value HR = 0. 97 (0. 78– 1. 20) P = 0. 77 HR = 0. 697 (0. 54– 0. 90) P = 0. 0059 OR = 2. 01 OR = 2. 22 (1. 41− 3. 47) P = 0. 0005 *RAS ascertainment rate: 80. 2%; †Primary endpoint; #WT 1. Stintzing S, et al. Ann Oncol 2014; 25(Suppl 5): v 1–v 41: abstract LBA 11 (and oral presentation); 2. Michl M, et al. J Clin Oncol 2014; 32(Suppl 5): abstract 3592 (and poster) KRAS exon 2 population.

§ PEAK – ETS ≥ 30% at Week 8 (1) WT RAS ETS ≥ 30% at Week 8, % 64% OR = 1. 99 (95% CI, 0. 99− 4. 10) P = 0. 052 45% Panitumumab + m. FOLFOX 6 (n = 80) Rivera F, et al. Eur J Cancer 2015; 51(Suppl 3): S 1‒S 810: abstract 2014 (and poster). Bevacizumab + m. FOLFOX 6 (n = 74)

§ PEAK – Depth of Response, Dp. R (2) WT RAS 65% P = 0. 0018† Dp. R, % 46% Panitumumab + m. FOLFOX 6 (n = 88) Rivera F, et al. Eur J Cancer 2015; 51(Suppl 3): S 1‒S 810: abstract 2014 (and poster). Bevacizumab + m. FOLFOX 6 (n = 81)

§ PEAK – % change from baseline in tumour load over time (3) WT RAS Mean change (95% CI) from baseline, % 0 − 20 − 40 − 60 − 80 Bevacizumab + m. FOLFOX 6 Panitumumab + m. FOLFOX 6 − 100 Pmab + m. FOLFOX 6 Bev + m. FOLFOX 6 0 8 16 24 32 Weeks 40 48 56 88 81 80 74 70 66 62 56 41 33 38 23 33 18 Rivera F, et al. Eur J Cancer 2015; 51(Suppl 3): S 1‒S 810: abstract 2014 (and poster). 53 44

Cremolini et al. Ann Oncol 2015; 26: 1188 -94

FOLFOXIRI + Bev FOLFIRI + Bev Impact: yes, but… Cremolini et al. Ann Oncol 2015; 26: 1188 -94

§ ETS & PFS, OS – overview of the literature (1) Heinemann V et al. Eur J Cancer 2015; 51: 1927 -36

§ ETS & PFS, OS – overview of the literature (2) Heinemann V et al. Eur J Cancer 2015; 51: 1927 -36

§ ETS, Dp. R & PFS, OS – randomised trials Heinemann V et al. Eur J Cancer 2015; 51: 1927 -36

§ ETS – summary • ETS reflects the velocity of response to therapy • It is measured at the earliest time point of CT evaluation • ETS indicates sensitivity to treatment • Addition of anti-EGFR agents to chemotherapy increases the number of patients with ETS and enhances the velocity of response Slide courtesy of V Heinemann

§ Dp. R – summary Δ OS Lethal tumour load Tumour size, % Δ OS 100% 70% SD PR • ETS is an early predictor of sensitivity to treatment PR • Dp. R correlates with OS Deepest response Time since start of treatment Adapted from: Stintzing S, et al. Ann Oncol 2014; 25 (Suppl 5): v 1–v 41: abstract LBA 11 (and oral presentation).

§ Additional response assessments – conclusions - ETS is associated with prolonged survival and is an early predictor of sensitivity to treatment in m. CRC - Based on ETS and Dp. R, anti-EGFR therapy may show a benefit over anti-VEGF therapy in 1 st-line WT RAS m. CRC - knowledge of the radiologist in standard setting - global population vs. individual decision

UZA DIAMOND DESIGN HARBOR I wish to thank The Multidisciplinary Oncology Team at Antwerp University Hospital