Diabetic Retinopathy Clinical Research Network Comparative Effectiveness of

Diabetic Retinopathy Clinical Research Network Comparative Effectiveness of Aflibercept, Bevacizumab, and Ranibizumab for DME Supported through a cooperative agreement from the National Eye Institute; National Institute of Diabetes and Digestive and Kidney Diseases; National Institutes of Health, Department of Health and Human Services EY 14231, EY 14229, EY 018817 1

Anti-VEGF Treatment Options for Diabetic Macular Edema Aflibercept 2. 0 -mg (EYLEA®) • • Better vision outcomes at 1 year for individuals with worse vision ~ U. S. cost per injection (2014): $1850 Bevacizumab 1. 25 -mg (Avastin®) • • • Not approved for DME by FDA ~ U. S. cost per injection (2014): $60 Typically repackaged into plastic syringes Ranibizumab 0. 3 -mg (Lucentis) • ~ U. S. cost per injection (2014): $1170

Relevance: Comparing Efficacy vs. Cost-effectiveness Comparing efficacy is most relevant to the doctor-patient relationship • For example, primary outcome of DRCR. net Protocol T • When initial visual acuity loss was mild prior to initiating anti-VEGF therapy for DME, there were no apparent differences, on average, in visual acuity outcomes at 1 year after initiating therapy with aflibercept, bevacizumab, or ranibizumab • At worse levels of initial visual acuity, aflibercept was more effective at improving vision at 1 year Comparing cost-effectiveness is most relevant to some payers and policymakers

Relevance: Efficacy vs. Costeffectiveness In 2010, about 2 billion dollars spent for anti-VEGF for ophthalmologic purposes – 1/6 th entire Medicare Part B drug budget Example: • July 2011: UK National Health Service (NHS) decided not to reimburse ranibizumab as a treatment of DME • April 2013: Novartis agrees to a discounted price to the UK government (details are “commercial in confidence”); NHS decides to reimburse ranibizumab Given increasing public health importance of DME and large differences in costs among different treatment alternatives for DME, a post-hoc costeffectiveness analysis of different anti-VEGF agents for DME was undertaken by DRCR. net

Cost-Effectiveness Plane Higher Cost Worse Health (QALYs) Better Health (QALYs) Lower Cost

Cost-Effectiveness Plane Incremental Cost Less Favorable COSTS MONEY, WORSENS HEALTH COSTS MONEY, IMPROVES HEALTH x ? DOMINATED SAVES MONEY, WORSENS HEALTH x/? A SAVES MONEY, IMPROVES HEALTH B √ DOMINANT More Favorable Incremental Benefit (QALYs Gained)

Cost-Effectiveness Ratio Incremental Cost-Effectiveness Ratio (ICER): Additional Costs you pay for B vs. A ($$$) Additional Health Benefits you get from B vs. A (QALYs) High ratios are “Bad”, low ratios are “Good” No explicit threshold, but generally. . . • <$50, 000/QALY considered good value • >$250, 000/QALY considered too expensive

Purpose Determine cost-effectiveness of anti-VEGF treatment options for patients with newlydiagnosed diabetic macular edema involving the center of the macula with visual acuity loss (20/32 or worse)

Costs “Societal” perspective Direct medical costs based on actual resource use from the trial and CMS allowables Costs of each intervention • Number of treatments per year from DCRC. net Protocol T • Tapering of treatments without average visual acuity loss as published from DRCR. net Protocol I Costs of managing side effects Costs of caring for patients who are blind from DME

Health-Related Quality of Life �Time spent in each health state is weighted by a quality multiplier to reflect quality of life 1 Perfect Health 0 Death

Sample Quality Adjustments. 92 – moderate angina. 87 – migraine. 82 – severe angina 1 Perfect Health. 90 – asymptomatic HIV. 84 – ulcer . 78 – atrioventricular blockage . 61 – mild schizophrenia . 53 – AIDS . 45 – severe clinical depression. 29 – severe chronic schizophrenia 0 . 40 – symptomatic, drugresistant prostate cancer. 07 – post-cardiac arrest, moderately impaired Death

Health-Related Quality of Life Note: Protocol T did not elicit participants’ quality-of-life directly But, other researchers studied large populations and found relationships between best-corrected visual acuity and quality-of-life weights used for this study • Brown o Tied to better-seeing eye o Ex: 20/20 BCVA = 0. 97 20/200 BCVA = 0. 66 • RESTORE Trial (ranibizumab vs. laser for DME) o Tied to eye treated with anti-VEGF Utility scores for side effects (MI, CVA, death) Determined ICER of each intervention relative to one another

Results – Utility, Year 1 Mean cumulative QALYs over 1 year A vs B A vs R R vs B B R Difference (P-value) All 0. 869 participants 0. 849 0. 857 0. 020 (0. 03) 0. 011 (0. 22) 0. 008 (0. 40) Visual acuity 20/50 or worse 0. 835 at baseline 0. 823 0. 829 0. 012 (0. 33) 0. 006 (0. 63) 0. 006 (0. 59) Visual acuity 20/32 to 20/40 0. 901 at baseline 0. 875 0. 884 0. 026 (0. 02) 0. 017 (0. 18) 0. 009 (0. 55) A A: aflibercept, B: bevacizumab, R: ranibizumab

Results – Costs, Year 1 $ 30, 000 Adverse events $ 25, 000 Laser photocoagulation therapy $ 20, 000 $ 15, 000 $ 10, 000 Non-study eye anti-VEGF injections $ 5, 000 Study eye anti. VEGF injections Al l p ar tic i W pan or ts se Be V tte A r V Al A l p ar tic i W pan or ts s Be e V tte A r V A $ - Aflibercept Bevacizumab Ranibizumab

Results – Costs, Year 1 $ 30, 000 Adverse events $ 25, 000 Laser photocoagulation therapy $ 20, 000 $ 15, 000 $ 10, 000 Non-study eye anti-VEGF injections $ 5, 000 Study eye anti. VEGF injections Al l p ar tic i W pan or ts se Be V tte A r V Al A l p ar tic i W pan or ts s Be e V tte A r V A $ - Aflibercept Bevacizumab Ranibizumab

Results – Costs, Year 1 $ 30, 000 Adverse events $ 25, 000 Laser photocoagulation therapy $ 20, 000 $ 15, 000 $ 10, 000 Non-study eye anti-VEGF injections $ 5, 000 Study eye anti. VEGF injections Al l p ar tic i W pan or ts se Be V tte A r V Al A l p ar tic i W pan or ts s Be e V tte A r V A $ - Aflibercept Bevacizumab Ranibizumab

Results – Costs, Year 1, All Participants $ 30, 000 $ 25, 000 Adverse events $ 20, 000 Laser treatment $ 15, 000 Nonstudyeye injections Study-eye injections $ 10, 000 $ 5, 000 $ Aflibercept Bevacizumab Ranibizumab

Results – Costs, Year 1, Initial Visual Acuity 20/50 Or Worse $ 30, 000 $ 25, 000 Adverse events $ 20, 000 Laser treatment $ 15, 000 Nonstudyeye injections Study-eye injections $ 10, 000 $ 5, 000 $ Aflibercept Bevacizumab Ranibizumab

Results – Costs, Year 1, Initial Visual Acuity 20/32 to 20/40 $ 30, 000 $ 25, 000 Adverse events $ 20, 000 Laser treatment $ 15, 000 $ 10, 000 Nonstudyeye injections $ 5, 000 $ Aflibercept Bevacizumab Ranibizumab

Results – Cost-effectiveness, 1 Yr Cost (2015 USD) All patients Utility (QALYs) Cost-effectiveness vs. bevacizumab ($/QALY)* Bevacizumab $4, 100 0. 849 – Ranibizumab $18, 600 0. 857 $1, 730, 000 Aflibercept $26, 100 Baseline visual acuity 20/50 or worse Bevacizumab $5, 000 0. 869 $1, 100, 000 0. 823 – Ranibizumab $20, 400 0. 829 $2, 450, 000 Aflibercept $28, 100 0. 835 $1, 870, 000 Baseline visual acuity 20/32 to 20/40 Bevacizumab $3, 200 0. 875 – Ranibizumab $16, 900 0. 884 $1, 500, 000 Aflibercept $24, 100 0. 901 $798, 000

Results – Cost-effectiveness Projections $ 3, 200, 000 ICER vs. bevacizumab, $/QALY Ranibizumab, all patients Aflibercept, all patients $ 1, 600, 000 $ 800, 000 $ 400, 000 $ 200, 000 0 5 10 15 20 Time horizon (years) 25 30

Results – Cost-effectiveness, 10 Yr Cost (2015 USD) All patients Utility (QALYs) Cost-effectiveness vs. bevacizumab ($/QALY)* Bevacizumab $39, 800 6. 80 – Ranibizumab $79, 400 6. 87 $603, 000 Aflibercept $102, 500 Baseline visual acuity 20/50 or worse Bevacizumab $40, 700 6. 98 $349, 000 6. 60 – Ranibizumab $81, 200 6. 65 $817, 000 Aflibercept $104, 500 6. 82 $287, 000 Baseline visual acuity 20/32 to 20/40 Bevacizumab $38, 900 7. 01 – Ranibizumab $77, 700 7. 09 $506, 000 Aflibercept $100, 600 7. 14 $474, 000

Injection Cost Thresholds, 1 Yr Current drug cost per dose (2015 USD) All patients Ranibizumab $1, 170 Aflibercept $1, 850 Baseline visual acuity 20/50 or worse Ranibizumab $1, 170 Aflibercept $1, 850 1 -year horizon

Injection Cost Thresholds, 1 Yr Current drug cost per dose (2015 USD) All patients 1 -year horizon Cost producing cost-effectiveness of $100, 000/QALY Ranibizumab $1, 170 $100 Aflibercept $1, 850 $240 Baseline visual acuity 20/50 or worse Ranibizumab $1, 170 $94 Aflibercept $1, 850 $250

Injection Cost Thresholds, 1 Yr Current drug cost per dose (2015 USD) All patients 1 -year horizon Cost producing Relative reduction cost-effectiveness from current cost of $100, 000/QALY Ranibizumab $1, 170 $100 91% Aflibercept $1, 850 $240 87% Baseline visual acuity 20/50 or worse Ranibizumab $1, 170 $94 92% Aflibercept $1, 850 $250 87%

Injection Cost Thresholds, 10 Yr Current drug cost per dose (2015 USD) All patients 10 -year horizon Cost producing Relative reduction cost-effectiveness from current cost of $100, 000/QALY Ranibizumab $1, 170 $230 80% Aflibercept $1, 850 $570 69% Baseline visual acuity 20/50 or worse Ranibizumab $1, 170 $190 84% Aflibercept $1, 850 $700 62%

Limitations �Extrapolating the findings of the DRCR. net trial beyond year one �Using best-corrected visual acuity alone as a surrogate for the impact of DME on healthrelated quality-of-life ◦ Visual needs may be varied from patient to patient ◦ Utilities assigned to best-corrected visual acuity alone may underestimate impact on health-related quality of life

Conclusions Over 1 -year study period, for individuals with worse initial visual acuity, incremental costeffectiveness ratios (ICERs) of aflibercept and ranibizumab when compared with bevacizumab were $1. 9 million and $2. 5 million per qualityadjusted life-year (QALY) Overall ICERs projected over 10 years when compared with bevacizumab were $350, 000/QALY for aflibercept and $600, 000/QALY for ranibizumab

Conclusions (continued) For treatment of eyes with worse initial visual acuity, anti-VEGF agent cost per injection would have to decrease by 62% for aflibercept or 84% for ranibizumab for those therapies to have ICERs of $100, 000/QALY relative to bevacizumab if evaluated over a 10 -year time horizon

Relevance Aflibercept 2. 0 -mg and ranibizumab 0. 3 -mg are unlikely to be cost-effective relative to bevacizumab for treatment of DME unless their prices decline substantially These results highlight the challenges when safety and efficacy results (of importance to patients and physicians providing their treatment) are at odds with cost-effectiveness results (of importance to some payers and policymakers)

Thank You on Behalf of Diabetic Retinopathy Clinical Research Network (DRCR. net) 52 clinical study sites Study participants who volunteered to participate in this trial DRCR. net Data and Safety Monitoring Committee DRCR. net investigators and staff 31