Short or long adjuvant treatment: can we use new trials to decide it? Roberto Labianca Direttore DIPO Bergamo Direttore Cancer Center Direttore USC Cure Palliative, Hospice e Terapia del Dolore

Background and Rationale Current standard of care for stage III colon cancer patients: 6 months of oxaliplatin-based adjuvant therapy with FOLFOX or CAPOX (XELOX) Oxaliplatin is associated with cumulative dose-dependent neurotoxicity Debilitating for many patients, both short- and long-term Nerve damage (e.g. numbness, tingling, pain) can persist long after discontinuation of therapy, sometimes permanently Dose reductions and early discontinuation of therapy are common Shorter duration treatment without loss of efficacy would be of benefit to patients and health care resources

12 weeks of ci 5-FU vs 6 months of 5-FU/LV RFS 5 Year (%) 12 wks Ci 5-FU 6 months 5-FU/LV HR (95% CI) p-value RFS 73.3 66.7 0.8 (0.62-1.04) p=0.10 OS 75.7 71.5 0.79 (0.61-1.03) p=0.08 OS Likelihood of 12 wks of ci 5-FU being inferior: P<0.005 N=801 Chau et al., Ann Oncol 2005

International Duration Evaluation of Adjuvant Chemotherapy (IDEA) Collaboration Academic collaboration of clinicians and statisticians from six randomized phase III trials (12 countries) SCOT (UK, Denmark, Spain, Australia, Sweden, New Zealand), TOSCA (Italy), Alliance/SWOG 80702 (US, Canada), IDEA France, ACHIEVE (Japan), HORG (Greece) Total of 12,834 patients with stage III disease included in analysis High number of patients needed to make sure with high confidence that we are not sacrificing efficacy of therapy for decreased toxicity

R Study Overview Objective: Non-inferiority design: 3 months FOLFOX* Reduce side-effects of therapy without giving up (too much) anti- cancer efficacy of therapy Non-inferiority design: As agreed upon by patient advocates and oncologists, shorter duration of therapy should not sacrifice more than 12% of benefit of adjuvant therapy In statistical terms: upper 95% confidence interval of Hazard Ratio (HR) of disease free survival (DFS) should not exceed 1.12 3 months Stage III Colon Cancer Patients R FOLFOX* or CAPOX* 6 months 12,834 patients *Investigator’s choice, no randomization

Non-inferiority Hypothesis Testing Statistical Conclusions Under Different Scenarios 3m TRT better 6m TRT better Superiority One-sided Type I Error Rate = 0.025 Power = 90% Require 3390 DFS Events Non-inferiority Not proven Inferiority Hazard Ratio 1.0 1.12 Non-Inferiority Margin Piaggio et al. JAMA 2012;308(24):2594-2604 TRT: treatment

Patient Characteristics by Study TOSCA (N=2402) SCOT (N=3983) IDEA France (N=2010) C80702 (N=2440) HORG (N=708) ACHIEVE (N=1291) Median Age, years 64 65 61 67 66 ECOG PS*   95% 71% 74% 82% 96% 1 5% 29% 25% 28% 18% 4% T Stage T1-2 13% 12% 8% 15% T3 75% 59% 70% 67% 78% 57% T4 14% N Stage N1 73% 69% N2 27% 31% 33% 26% Chemotherapy CAPOX 35% 10% 0% 58% FOLFOX 65% 90% 100% 42% Median follow-up time, m 62 37 51 35 48 T4 12% 29% 18% 15% 14% 28% Median follow-up time, m 62 37 51 35 48 CAPOX 35% 67% 10% 0% 58% 75% This table shows patient characteristics for each study. Of note, two studies enrolled a higher percentage of patients with T4 disease More importantly, the percentage of patients that received CAPOX varied from 0% to 75% across individual studies. All studies reached at least 35m of follow-up. Beyond that, the median follow-up varies across studies. *1% of PS 2 in IDEA France and C80702 trials

Adverse Events FOLFOX CAPOX Overall G2 G3-4 32% 38% 57% 41% 24% 48% 3m Arm 6m Arm p-value1 Overall G2 G3-4 32% 38%  57%  <.0001   41% 24% 48% 37% Neurotoxicity 14% 3%  16% 12% 3% 36% 9%  Diarrhea 11% 5% 13% 7% 10% 0.0117 Again, as expected, adverse event rates are lower in 3m arm compared to the 6m arm. Especially, the rate of grade 2 or higher neurotoxicity in 6m arm is almost 3 times of that in the 3m arm. 1Chi-squared test for trend; Total of 19 grade 5 events; Adverse events only collected on first 617 patients enrolled to SCOT trial

Primary Outcomes Analysis Duration 100 3 Months 90 90 6 Months 80 80 70 70 60 60 Duration 3-yr DFS 3m 74.6 % 6m 75.5 % 3-yr DFS diff. = -0.9% , 95% CI, (-2.4 to 0.6%) Percent Without Event 50 50 DFS HR = 1.07 95% CI, 1.00 to 1.15 40 40 30 30 20 20 10 10 1 1 2 2 3 3 4 4 5 5 6 6 Years from Randomization Years from Randomization N Patients At risk 6424 5446 4464 3000 1609 826 321 6410 5530 4477 3065 1679 873 334

Non-inferiority Hypothesis Testing Statistical Conclusions Under Different Scenarios 3m TRT better 6m TRT better Superiority One-sided Type I Error Rate = 0.025 Power = 90% Require 3390 DFS Events Non-inferiority Not proven Inferiority Hazard Ratio 1.0 1.12 Non-Inferiority Margin Piaggio et al. JAMA 2012;308(24):2594-2604 TRT: treatment

Primary DFS Analysis (mITT), cont. Statistical Conclusions 3m TRT better 6m TRT better DFS HR = 1.07 95% CI, 1.00 to 1.15 Not proven 1.0 1.12 Hazard Ratio Non-Inferiority Margin Recall, in order for the 3m treatment duration to be designated as non-inferior, the upper bound of the CI for the DFS HR, needed to be 1.12 or less. The results showed that the DFS HR comparing 3m and 6m treatment is 1.07. And most importantly, the upper bound of the CI was 1.15, which is greater than pre-specified NI margin of 1.12. In other words, the non-inferiority of 3m of adjuvant treatment compared to 6m of treatment was NOT established in overall stage III colon cancer. Pause However, when the primary results are inconclusive and we have access to a very large sample size, as in the current study, it is important to examine whether the result are different for pre-planned subgroups. Let’s move onto subgroup analysis next. TRT: treatment

Analysis by Risk Groups and Regimens Large difference in overall prognosis observed between (T1-3 N1) and (T4 and/or N2) cancers 3 year DFS ∆ 20% (~80% vs 60%) Preplanned subgroup analysis for T and N Analysis of 3m vs 6m adjuvant therapy by risk groups Two different adjuvant regimens used, FOLFOX and CAPOX Preplanned analysis of 3m vs 6m based on regimen

CAPOX/FOLFOX combined 3 yr DFS rate (%) and HR by regimen and risk group Regimen CAPOX FOLFOX CAPOX/FOLFOX combined 3 yr DFS, % (95% CI) HR (95% CI) 3 m 6 m Risk group Low-risk (T1-3 N1) ~60% High-risk (T4 and / or N2) ~40% Risk groups combined Non-inferior Not proven Inferior

CAPOX/FOLFOX combined 3 yr DFS rate (%) and HR by regimen and risk group Regimen CAPOX FOLFOX CAPOX/FOLFOX combined 3 yr DFS, % (95% CI) HR (95% CI) 3 m 6 m Risk group Low-risk (T1-3 N1) ~60% 85.0 (83.1-86.9) 83.1 (81.1-85.2) 0.85 (0.71-1.01) High-risk (T4 and / or N2) ~40% Risk groups combined Non-inferior Not proven Inferior

CAPOX/FOLFOX combined 3 yr DFS rate (%) and HR by regimen and risk group Regimen CAPOX FOLFOX CAPOX/FOLFOX combined 3 yr DFS, % (95% CI) HR (95% CI) 3 m 6 m Risk group Low-risk (T1-3 N1) ~60% 85.0 (83.1-86.9) 83.1 (81.1-85.2) 0.85 (0.71-1.01) High-risk (T4 and / or N2) ~40% 61.5 (58.9-64.1) 64.7 (62.2-67.3) 1.20 (1.07-1.35) Risk groups combined Non-inferior Not proven Inferior

CAPOX/FOLFOX combined 3 yr DFS rate (%) and HR by regimen and risk group Regimen CAPOX FOLFOX CAPOX/FOLFOX combined 3 yr DFS, % (95% CI) HR (95% CI) 3 m 6 m Risk group Low-risk (T1-3 N1) ~60% 85.0 (83.1-86.9) 83.1 (81.1-85.2) 0.85 (0.71-1.01) High-risk (T4 and / or N2) ~40% 64.1 (61.3-67.1) 64.0 (61.2-67.0) 1.02 (0.89-1.17) 61.5 (58.9-64.1) 64.7 (62.2-67.3) 1.20 (1.07-1.35) 62.7 (60.8-64.4) 64.4 (62.6-66.4) 1.12 (1.03-1.23) Risk groups combined 75.9 (74.2-77.6) 74.8 (73.1-76.6) 0.95 (0.85-1.06) Non-inferior Not proven Inferior

CAPOX/FOLFOX combined 3 yr DFS rate (%) and HR by regimen and risk group Regimen CAPOX FOLFOX CAPOX/FOLFOX combined 3 yr DFS, % (95% CI) HR (95% CI) 3 m 6 m Risk group Low-risk (T1-3 N1) ~60% 85.0 (83.1-86.9) 83.1 (81.1-85.2) 0.85 (0.71-1.01) 81.9 (80.2-83.6) 83.5 (81.9-85.1) 1.10 (0.96-1.26) 83.1 (81.8-84.4) 83.3 (82.1-84.6) 1.01 (0.90-1.12) High-risk (T4 and / or N2) ~40% 64.1 (61.3-67.1) 64.0 (61.2-67.0) 1.02 (0.89-1.17) 61.5 (58.9-64.1) 64.7 (62.2-67.3) 1.20 (1.07-1.35) 62.7 (60.8-64.4) 64.4 (62.6-66.4) 1.12 (1.03-1.23) Risk groups combined 75.9 (74.2-77.6) 74.8 (73.1-76.6) 0.95 (0.85-1.06) 73.6 (72.2-75.1) 76.0 (74.6-77.5) 1.16 (1.06-1.26) Non-inferior Not proven Inferior

CAPOX/FOLFOX combined P-value interaction test: 3 yr DFS rate (%) and HR by regimen and risk group Regimen CAPOX FOLFOX CAPOX/FOLFOX combined 3 yr DFS, % (95% CI) HR (95% CI) 3 m 6 m Risk group Low-risk (T1-3 N1) ~60% 85.0 (83.1-86.9) 83.1 (81.1-85.2) 0.85 (0.71-1.01) 81.9 (80.2-83.6) 83.5 (81.9-85.1) 1.10 (0.96-1.26) 83.1 (81.8-84.4) 83.3 (82.1-84.6) 1.01 (0.90-1.12) High-risk (T4 and / or N2) ~40% 64.1 (61.3-67.1) 64.0 (61.2-67.0) 1.02 (0.89-1.17) 61.5 (58.9-64.1) 64.7 (62.2-67.3) 1.20 (1.07-1.35) 62.7 (60.8-64.4) 64.4 (62.6-66.4) 1.12 (1.03-1.23) Risk groups combined 75.9 (74.2-77.6) 74.8 (73.1-76.6) 0.95 (0.85-1.06) 73.6 (72.2-75.1) 76.0 (74.6-77.5) 1.16 (1.06-1.26) P-value interaction test: Regimen: 0.0061 Risk group: 0.11 Non-inferior Not proven Inferior

IDEA Recommendations Regimen CAPOX FOLFOX 3 months (3-)6 months Risk group Low-risk (T1-3 N1) ~60% 3 months (3-)6 months High-risk (T4 and/or N2) ~40% 3(-6) months 6 months

Conclusions IDEA data provide a framework for discussions on risks and benefits of individualized adjuvant therapy approaches Shorter duration of therapy associated with remarkable reduction in (neuro)toxicity For patients treated with CAPOX, 3 months was as good as 6 months, particularly in the low-risk population. For patients treated with FOLFOX, 6 months treatment added extra benefit in terms of DFS, particularly in the high-risk population

Acknowledgement All IDEA group investigators Patients, families, and caregivers who participated in each of these studies Greg Yothers, PhD Academic funding support: U10CA180821, U10CA180882, U10CA180888, U10CA180835, INCA, PHRC2009, EME 09/800/34, HTA 14/140/84, CRUK C1348/A15960, JFMC