Insights & lessons from the ODYSSEY Outcomes trial: Which patients benefit most? Ph.Gabriel Steg RHU iVASC Hôpital Bichat, Assistance Publique – Hôpitaux de Paris, Université de Paris, INSERM U-1148, Paris, France, FACT: French Alliance for Cardiovascular clinical Trials @gabrielsteg
Disclosures - Research grants : Amarin, Bayer, Sanofi, and Servier - Clinical Trial Contract (Steering committee or CEC) : Amarin, AstraZeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Novartis, Pfizer, Sanofi, Servier - Consulting or speaking : Amgen, Novo-Nordisk, Regeneron
ODYSSEY OUTCOMES ACS patients (hospitalized 1 to 12 months before randomization) Run- in period of 2−16 weeks on high - intensity† or maximum -tolerated dose of atorvastatin or rosuvastatin At least one lipid entry criterion met Randomization Placebo SC Q2W Alirocumab SC Q2W Patient and investigators remained blinded to treatment and lipid levels for the entire duration of the study † High-intensity statin defined as atorvastatin 40 or 80 mg daily or rosuvastatin 20 or 40 mg daily. ACS, acute coronary syndromes; Q2W, every two weeks; SC, subcutaneous. Schwartz GG, Steg PG, et al. N Engl J Med 2018
A Target Range for LDL-C Undesirably high Target baseline range We attempted to range maximize the number of Acceptable range Below target patients in the target range and minimize the number below target by Alirocumab blindly titrating alirocumab (75 or 150 mg SC Q2W) or blindly switching to placebo. 0 15 25 50 70 LDL-C (mg/dL) Schwartz GG, et al. Am Heart J 2014;168:682-689.e1.
Primary Efficacy Endpoint: MACE CHD death, non-fatal MI, ischemic stroke, or UA requiring hospitalization (95% CI: 13.5 – 15.6) 14.5% (95% CI: 11.5 – 13.5) 12.5% To prevent one primary end point event 49 patients (95% CI: 28 – 164) would need to be treated for 4 years * HR 0.85 (95% CI 0.78, 0.93) P=0.0003 *For patients with baseline LDL-C ≥ 100 mg/dL , 16 patients (95% CI, 11 to 34) would need to be treated for 4 years Schwartz GG, Steg PG, et al. N Engl J Med 2018
Primary Efficacy and Components Alirocumab Placebo Log-rank Endpoint, n (%) HR (95% CI) (N=9462) (N=9462) P-value MACE 903 (9.5) 1052 (11.1) 0.85 (0.78, 0.93) 0.0003 CHD death 205 (2.2) 222 (2.3) 0.92 (0.76, 1.11) 0.38 Non-fatal MI 626 (6.6) 722 (7.6) 0.86 (0.77, 0.96) 0.006 Ischemic stroke 111 (1.2) 152 (1.6) 0.73 (0.57, 0.93) 0.01 Unstable angina 37 (0.4) 60 (0.6) 0.61 (0.41, 0.92) 0.02
All-Cause Death HR 0.85 (95% CI 0.73, 0.98) P=0.026* *Nominal P-value Steg PG et al. Circulation 2019
Effect of Alirocumab on All-Cause, CV, and Non-CV Death Incidence 7 Alirocumab Placebo Log-rank Placebo Endpoint HR (95% CI)* n (%) n (%) P -value* All-Cause, CV, or Non-CV Death (%) Alirocumab 6 † All-Cause Death 334 (3.5) 392 (4.1) 0.85 (0.73, 0.98) .03 5 All-cause 4 CV Death 240 (2.5) 271 (2.9) 0.88 (0.74, 1.05) .15 3 CV 2 Non-CV Death 94 (1.0) 121 (1.3) 0.77 (0.59, 1.01) .06 Non-CV 1 0.6 0.85 1.0 1.18 0 *Stratified by geographic region Alirocumab Placebo 0 1 2 3 4 † Nominal P -value Better Better Years Since Randomization Number at risk Placebo 9462 9219 8888 3898 737 Alirocumab 9462 9217 8919 3946 746 Steg PG et al. Circulation 2019
All-cause Death According to Achieved LDL-C in the Alirocumab Group at Month 4 Steg PG et al. Circulation 2019
Safety Alirocumab Placebo Event (N=9451) (N=9443) Diabetes worsening or diabetic complications: pts 506/2688 (18.8) 583/2747 (21.2) w/DM at baseline , n/N (%) New onset diabetes; pts w/o DM at baseline , n/N (%) 648/6763 (9.6) 676/6696 (10.1) General allergic reaction, n (%) 748 (7.9) 736 (7.8) Hepatic disorder, n (%) 500 (5.3) 534 (5.7) Local injection site reaction, n (%)* 360 (3.8) 203 (2.1) Neurocognitive disorder, n (%) 143 (1.5) 167 (1.8) Cataracts, n (%) 120 (1.3) 134 (1.4) Hemorrhagic stroke, n (%) 9 (<0.1) 16 (0.2) Schwartz GG, Steg PG, et al. N Engl J Med 2018 *HR vs placebo 1.82 (95% CI 1.54, 2.17)
Mean Cumulative Functions for First and Total Nonfatal CV Events NNT (95% CI) for 4 years: • 18 (11, 53) for total events • Placebo: Total Nonfatal CV 44 (26, 129) for first events Alirocumab: Total Nonfatal CV Placebo: First Nonfatal CV Alirocumab: First Nonfatal CV Szarek M et al. JACC 2019
Median on-treatment lipids (and change from baseline) similar across glucometabolic categories in the ODYSSEY OUTCOMES trial Normoglycemia Prediabetes Diabetes 200 Placebo Median (Q1, Q3), mg/dL Alirocumab 160 Placebo 120 Placebo Alirocumab 0% +1% 0% -2% 80 +1% 0% -16% -15% -14% Alirocumab +1% Placebo +1% Alirocumab 0% 40 -54% -54% +8% -54% +8% +7% +3% +3% +3% -64% -64% -65% 0 LDL-C Non-HDL-C HDL-C Triglyceride Measurements at Week 16 of assigned treatment; median change from baseline indicated below bars. Intention-to-treat analysis Ray KK et al. Lancet Diabetes Endoc 2019
Similar relative, but greater absolute risk reduction with alirocumab in patients with diabetes Relative risk reduction Interaction Absolute risk reduction P=0.98 Interaction P=0.002 MACE Incidence Alirocumab Placebo Absolute risk reduction AR ARR (95% CI) n/N (%) n/N (%) Subgroup HR (95% CI) (95% CI) Overall 903/9462 (9.5) 1052/9462 (11.1) 0.85 (0.78, 0.93) 1.6% 1.6% (0.7%, 2.4%) Normoglycemia 192/2639 (7.3) 220/2595 (8.5) 0.85 (0.70, 1.03) 1.2% 1.2% (-0.3%, 2.7%) Prediabetes 331/4130 (8.0) 380/4116 (9.2) 0.86 (0.74, 1.00) 1.2% 1.2% (0%, 2.4%) Diabetes 380/2693 (14.1) 452/2751 (16.4) 0.84 (0.74, 0.97) 2.3% 2.3% (0.4%, 4.2%) 0.75 0.85 1.0 3.2% 1.6% 0% Ray KK et al. Lancet Diabetes Endoc 2019 Alirocumab Placebo Placebo Alirocumab Placebo Better Better Better Better Better
No increase in HbA1c, fasting glucose, or new-onset diabetes with alirocumab among patients without diabetes at baseline F 5.9 6.0 18 HbA1c Fasting glucose New onset diabetes 5.8 15 5.9 All Patients Analysis method for A1c and fasting Mean (95% CI) Fasting Glucose, mmol/L All Patients All Patients Without Diabetes glucose: repeated-measures mixed effects model; random effects = Without Diabetes Without Diabetes p=0.84 slope, intercept; fixed effects = p=0.0008 5.7 HR (95% CI) = Mean (95% CI) HbA1c, % treatment, baseline value, and time. Incidence (95% CI), % 12 1.00 (0.89-1.11) 5.8 Only post-randomization values prior to initiation of diabetes medication were included in the analysis. 5.6 9 Without diabetes = prediabetes or normoglycemia. 5.7 5.5 6 5.6 5.4 3 Ray KK et al. Lancet 5.3 5.5 Diabetes Endoc 2019 0 0 0 Alirocumab Placebo Alirocumab Placebo Alirocumab Placebo o
MACE: one, two or three vascular beds Jukema et al JACC 2019
Pr Primary imary En Endpoint point Accor ccording ding to o Pr Prio ior r CABG BG St Status tus *Index CABG is CABG between the index ACS event and randomization (including 44 patients with prior CABG). †Prior CABG is CABG prior to the index ACS event. Goodman SG et al. J ACC (in press)
Primary Endpoint and All-Cause Death According to Prior CABG Status *Index CABG is CABG between the index ACS event and randomization (including 44 patients with prior CABG). †Prior CABG is CABG prior to the index ACS event. Goodman SG et al. J ACC (in press) HRs reflect stratification by geographic region. ACS indicates acute coronary syndrome; ARR, absolute risk reduction; CABG, coronary artery bypass graft; CI, confidence interval; and HR, hazard ratio.
First occurrence of type 1 or 2 MI, and effects of alirocumab over time P =0.032 Placebo P =0.025 White HD et al. Eur Heart J 2019
Baseline Lp(a) Predicts MACE Risk in the Placebo Group Bittner et al. ACC’19
Baseline Lp(a) Predicts Absolute Change in Lp(a), but not LDL-C Bittner et al. ACC’19 Alirocumab group
Change in Lp(a) Predicts MACE, Independent of LDL-C corr Change HR (95% CI) per Model Model Adjustments Parameter 1 mg/dl decrease p-value 1 Baseline Lp(a) Lp(a) 0.993 (0.989, 0.998) 0.0027 Baseline Lp(a), Baseline LDL-C corr , Lp(a) 0.994 (0.990, 0.999) 0.0081 2 Change from Baseline to Month 4 LDL-C corr 0.996 (0.994, 0.998) 0.0002 in LDL-C corr Bittner et al. ACC’19 Changes in lipoproteins measured between baseline and Month 4
Implications of Hazard Ratios: Some Examples Lp(a) Reduction HR* RRR for MACE (mg/dL) 1 0.994 0.6% 0.994 5 = 0.970 5 3.0% 0.994 10 = 0.942 10 5.8% 0.994 15 = 0.914 15 8.6% 0.994 20 = 0.890 20 11.0% * independent of baseline Lp(a), baseline LDL-C corr and change in LDL-C corr 5 mg/dL reduction = median; 15 mg/dL reduction = 75 th percentile Bittner et al. ACC’19
Proportion of MACE Reduction Attributable to Changes in Lp(a) and Corrected LDL-C 100 89% 73% 96% 80 Percent LDL-C corr 60 40 Lp(a) 20 27% 4% 11% 0 25th 50th 75th Baseline Lp(a) percentile Baseline Lp(a) in mg/dL 6.7 21.2 59.6 Bittner et al. ACC’19 From model with baseline and change in Lp(a), baseline and change in LDL-C corr (Model 2)
High-risk post ACS patients likely to derive greater absolute benefit from PCSK9 inhibition • Patients with LDL > 100 mg/dL (and HeFH patients) • Diabetic patients • Polyvascular patients • Post CABG patients • High Lp(a) patients
Recommend
More recommend
