STUDY SUBJECTS

Major Inclusion criteria:

• Treatment-naive NPDR with no diabetic macular edema (DME)
• Best-corrected visual acuity (BCVA) of 20/40 (Snellen equivalent) or better

RANDOMIZATION SCHEME AND INTERVENTIONS

Randomized (stratified by baseline DRSS level (47; 53)) 1:1:1 to

• Aflibercept 2 mg q1m x 3, then q8w x 1, followed by q16 weeks (aflibercept 2q16 group)
• Aflibercept 2 mg q1m x 5 followed by q8w during year 1, followed by flexible pro re nata (PRN) dosing in year 2. Re-treatment criteria for PRN therapy was DRSS level worse than 35
• Sham injections (control group)

Primary Endpoint

• Proportion of eyes with a 2-step or greater improvement in DRSS level from baseline to week 24 and week 52

Secondary Endpoints

• Proportion of eyes that developed vision-threatening complications or center-involved diabetic macular edema (CI-DME) at week 52 and week 100
• Proportion of eyes with a 2-step or greater worsening in DRSS level from baseline to week 52 and week 100

Results

Study population:

• 402 participants
• 88.3% completion of week 52 and 79.6% completion of week 100, with higher completion rates in the aflibercept groups vs the control group at weeks 52 and 100

Primary Outcome: Improvement in DRSS level from baseline (2-step improvement)

• At week 24
  • combined aflibercept groups: 58.4%
  • controls: 6.0%
• At week 52
  • aflibercept2q16: 65.2%
  • aflibercept 2q8/PRN: 79.9%
  • control: 15.05%
• At week 100
  • aflibercept 2q16: 62.2%
  • aflibercept 2q8/PRN: 50.0%
  • controls: 12.8% among 53 eyes in the control group that received rescue treatment for CI-DME, 21 eyes (39.6%) had a 2-step or greater improvement in DRSS level from baseline

Secondary Outcomes

• Significantly lower rates of DRSS level worsening from baseline at week 52 in aflibercept groups
  • aflibercept 2q16: 1.6%
  • aflibercept 2q8/PRN group: 0%
  • control group: 11.9% at week 52.
• Significantly lower rates of DRSS level worsening from baseline at week 100 in aflibercept groups
  • aflibercept 2q16: 4.5%
  • aflibercept 2q8/PRN: 2.4%
  • control: 20.2%
• Significantly fewer eyes developed vision-threatening complications at 52 and 100 weeks respectively with aflibercept
  • aflibercept 2q16: 3.7%, 8.1%
  • aflibercept 2q8/PRN: 3.0%, 6.0%
  • control: 20.3%, 27.1%
• Significantly fewer eyes in the aflibercept groups developed CI-DME at 52 and 100 weeks respectively
  • aflibercept 2q16: 6.7%, 10.4%
  • aflibercept 2q8/PRN: 8.2%, 13.4%
  • control: 25.6%, 33.1%

Rescue Treatment

• The proportion of participants who underwent PRP or vitrectomy was lower in the aflibercept groups compared with the control group

Changes in Central Subfield Thickness (CST) and BCVA

• At week 52 and week 100 respectively, changes in CST were
  • -18.9 μm, -18.6 μm in the aflibercept 2q16 group
  • -24.9 μm, -15.2 μm in the aflibercept 2q8/PRN group
  • +5.3 μm, +10.3 μm in the control group
• At week 52 and week 100 respectively, the mean (SD) area under the curve (AUC) for the change in BCVA was
  • +1.7 letters, +1.5 letters in the aflibercept 2q16 group
  • +1.3 letters, +0.8 letters in the aflibercept 2q8/PRN group
  • +0.5 letters, +0.6 letters in the control group

Flexible (PRN) Dosing At Year 2

• 4 eyes (9.8%) that received no injections developed a vision-threatening complication compared to 0 eyes (0%) that received 1 or more injection
• 4 eyes (9.8%) that received no injections and 3 eyes (7.1%) that received 1 or 2 injections developed CI-DME compared to 0 eyes (0%) that received 3 or more injections

Adverse Events

• Ocular adverse events occurred in more than 7.5% of eyes in the control and aflibercept groups
• No cases of occlusive retinal vasculitis or endophthalmitis were observed
• Nonocular serious adverse events were similar across treatment groups

Conclusions

• Treatment of severe NPDR with intravitreal aflibercept injections improved the severity of retinopathy and reduced the risk of progression to CI-DME and vision-threatening complications
• No difference in BCVA at the 2-year mark
• Treatment at a fixed interval may be an optimal strategy compared with reactive PRN dosing