Diabetic Retinopathy Clinical Research Network Changes in Diabetic

Diabetic Retinopathy Clinical Research Network Changes in Diabetic Retinopathy Severity when Treating Diabetic Macula Edema with Ranibizumab: DRCR. net Protocol I 5 -year Report 1

Background Ø Phase 3 DME trials have shown eyes assigned to ranibizumab or aflibercept to manage DME have favorable changes in diabetic retinopathy severity relative to controls when followed for 2 -3 years* Ø More eyes improve Ø Fewer eyes worsen Ø Long term follow-up is needed to determine: Ø If these outcomes are maintained … Ø Particularly when using a treatment regimen in which the exposure to anti-VEGF therapy decreases over time. * VISTA&VIVID, DRCR. net Protocol I 3 -Year data, RISE&RIDE, Protocol T. 2

3 -Year Retinopathy Worsening: Protocol I* Eyes with Baseline NPDR Cumulative Probability 50% 40% A: Laser B: Ranibizumab + Laser C: Ranibizumab + Def D: Triamcinolone + Laser 37% 30% 23% 20% 18% 10% 7% 0% 1 Year *JAMA Ophthalmology 2013 2 Years 3

3 -Year Retinopathy Worsening: Protocol I* Eyes with Baseline PDR Cumulative Probability 50% 40% A: Laser B: Ranibizumab + Laser C: Ranibizumab + Def D: Triamcinolone + Laser 40% 30% 21% 20% 18% 10% 0% 12% 1 Year *JAMA Ophthalmology 2013 2 Years 3 Years 4

Purpose To report DRCR. net Protocol I 5 -year rates of retinopathy improvement or worsening: • among eyes assigned to ranibizumab • in the context of progressive reduction in exposure to ranibizumab therapy 5

Methods Ø Rates of improvement and worsening were evaluated: • In the ranibizumab treatment groups o Ranibizumab + deferred laser COMBINED INTO o Ranibizumab + prompt laser ONE GROUP • Within each DR severity subgroup*: o NPDR group (N=235) • Level 10 through 53, i. e. , very severe NPDR or less o PDR group (N=111) • Level 60 or greater, i. e. , PRP, inactive PDR through advanced PDR *Based on Reading Center grading of baseline fundus photos 6

Improvement of Diabetic Retinopathy between Baseline and each Annual Visit* Limited to annual visits that required 7 field photographs 7

Criteria for Improvement Did NOT receive PRP During follow-up on annual photos Improvement on annual fundus photos* Improvement Did NOT develop NVI or NVA or NVG of Diabetic Retinopathy Between Did NOT baseline receive vitrectomy and or injections for PDR annual follow-up Did NOT develop vitreous hemorrhage Did NOT develop retinal detachment 8 *Based on Reading Center grading of annual fundus photos, excludes <20 (MA)

Criteria for Improvement Improved 2 or more levels Did NOT on receive the ETDRS retinopathy PRP scale* During follow-up on annual photos Improvement on annual fundus photos* Did NOT develop NVI or NVA or NVG Between (applies to eyes with NPDR Did NOTor baseline receive 2 PDR with potential to improve vitrectomy or and or more levels) injections for PDR annual follow-up Improvement of Diabetic Improved from active PDR Retinopathy to Inactive (if PRP at baseline) Did NOT or No PDR develop vitreous (if no PRP at baseline) * hemorrhage Did NOT develop retinal detachment 9 *Based on Reading Center grading of annual fundus photos, excludes <20 (MA)

Percentage of Eyes with Improvement of Diabetic Retinopathy With Ranibizumab Injections* NPDR : P for trend=0. 78 50% 38% 40% 30% PDR : P for trend=0. 30 29% 35% 28% 32% 25% 23% 20% 10% 0% Year 1 N 191 40 Year 3 170 34 Year 4 148 28 Year 5 137 *among participants who were eligible to improve and had gradable fundus photograph at the specified annual visit 26 10

Percentage of Eyes with Improvement of Diabetic Retinopathy With Ranibizumab Injections* NPDR : P for trend=0. 78 50% 38% 40% 30% PDR : P for trend=0. 30 29% 35% 28% 32% 25% 23% 20% 10% 0% Year 1 Year 3 Year 4 Year 5 N 191 40 170 34 148 28 137 26 Median # of Injections 8 8 1 0 0 0, 3 0, 2 IQR 6, 10 7, 10 0, 4 eligible to 0, 3 improve 0, 2 and *among participants 0, 4 who were had gradable fundus photograph at the specified annual visit 11

Percentage of Eyes with Improvement of Diabetic Retinopathy With Ranibizumab Injections* NPDR : P for trend=0. 78 50% 38% 40% 30% PDR : P for trend=0. 30 29% 35% 28% 25% 23% 20% 10% 0% Year 1 Year 3 Year 4 Year 5 N 191 40 170 34 148 28 26 Median # of Injections 8 8 1 0 0 IQR 6, 10 7, 10 0, 4 eligible to 0, 3 improve 0, 2 and *among participants 0, 4 who were had gradable fundus photograph at the specified annual visit 0, 2 12

Improvement of Diabetic Retinopathy With Ranibizumab Injections* Percentage of Eyes with Improvement Moderate NPDR or less vs Moderate-severe/Severe NPDR P=0. 002 50% 38% 40% 30% 42% 29% 35% 28% 25% 20% 23% 15% 10% 0% Year 1 Year 3 Year 4 Year 5 N 191 40 170 34 148 28 52 85 26 Median # of Injections 8 8 1 0 0 0 1 0 0 IQR 6, 10 7, 10 0, 4 eligible to 0, 3 improve 0, 2 and 0, 3 0, 4 0, 2 *among participants 0, 4 who were had gradable fundus photograph at the specified annual visit 13

Relationship between injection number and improvement Number of injections between Baseline and the Specified Visit Improved Not Improved P-value 1 -Year Median (IQR) N=70 10 (7, 11) N=161 7 (6, 9) <0. 001 3 -Year Median (IQR) N=60 14 (8, 19) N=144 11 (8, 16) 0. 02 4 -Year Median (IQR) N=49 15 (10, 23) N=127 13 (8, 19) 0. 02 Eyes given more injections were N=113 more likely 5 -Year N=50 all annual 15 visits Median (IQR) 16 (11, 30) (10, 22) to improve at 0. 006 14 *among participants who were eligible to improve and had gradable fundus photograph at each annual visit

mprovement Status at 1 and 3 Years* Diabetic Retinopathy Status Relative to Baseline 47% of eyes with improvement at Label 1 -Year 3 -Year 1 year sustained improvement at Improvement Improved 3 years sustained N 25 Improvement not maintained Improved Not improved 28 Late improvement Not improved Improved 26 No improvement Not improved 107 15 *among participants who were eligible to improve and had gradable fundus photograph at each annual visit

Worsening of Diabetic Retinopathy through 5 Years* * subjects that did not complete 5 -year visit AND did not meet worsening definition were censored at the last completed visit 16

Criteria for Worsening Between baseline and follow-up Received vitrectomy or injections Developed retinal detachment Developed NVI or NVA or NVG Developed vitreous hemorrhage Only applies to eyes without PDR at baseline Worsened 2 or more levels on the ETDRS retinopathy scale* During follow-up on annual photos Progressed from No PDR to PDR* for PDR Received PRP Worsening of Diabetic Retinopathy Progressed from ≤HRPDR to advanced PDR* 17

Cumulative Probability of Retinopathy Worsening With Ranibizumab Injections* Cumulative probabilities of worsening 40% PDR vs NPDR; P=0. 01 30% NPDR 20% 8% 10% 23% 19% 16% 31% 15% 12% 18% 5% 2% 0% 0 # at risk 1 218 2 96 195 *Calculated Time in Years 85 167 3 76 using the lifetable method 4 145 56 5 127 45 18

Cumulative Probability of Retinopathy Worsening With Ranibizumab Injections* Cumulative probabilities of worsening 40% Rate of worsening did not change with time P > 0. 05 30% NPDR 20% 8% 10% 23% 19% 16% 31% 15% 12% 18% 5% 2% 0% 0 1 2 Time in Years 3 4 5 # at risk 218 96 195 85 167 76 145 56 127 45 Median # of Injections 8 10 2 3 1 1 0 0 0, 5 0, 8 0, 4 0, 5 0, 3 0, 4 0, 3 IQR 6, 10 7, 12 *Calculated using the lifetable method 19

Cumulative Probability of Retinopathy Worsening With Ranibizumab Injections* Cumulative probabilities of worsening 40% 31% Moderate NPDR or better 30% Moderate-severe/Severe NPDR 23% PDR 20% 19% 16% 8% 10% 4% 7% 1 2 0% 0 # at risk 87 131 96 20% 8% 5% Time in Years 82 114 85 *Calculated 18% 3 77 91 76 using the lifetable method 24% 11% 4 5 62 83 56 20

Top 3 Events Triggering Worsening Change on annual 10 eyes met more than photographs one worsening criteria and 16 eyes developed NV Vitreous hemorrhage at some point NPDR N=36 PDR N=29 36% 7% 33% 55% Panretinal 17% 21% 6 eyes met more than photocoagulation one worsening criteria at some point 21

Strengths and Limitations Strengths Ø Prospective Ø Long term follow-up Limitations Ø Annual visits without gradable fundus photographs 14%/19%/31% at 1/3/5 -yr visits* Ø Limited number of eyes with baseline PDR Ø Conditions other than PDR may have led to vitreous hemorrhage which may have overestimated the worsening rates in the PDR group 22 *among living participants

Conclusion Ø Individuals managed with anti-VEGF therapy for DME may have simultaneous favorable alterations in the evolution of their retinopathy Ø In this exploratory analysis, Improvement and Worsening rates did not appear to be altered by the marked reduction in exposure to ranibizumab during the later years of follow-up • However, individual eyes may not maintain improvement, and worsening does occur, highlighting need for continued surveillance of DR level. 23

Thank You 24