A Cost-Effectiveness Analysis of Intravitreal Aflibercept for the Prevention of Progressive Diabetic Retinopathy

This was a cost-effectiveness analysis based on published data from two recent randomized control trials studying the early treatment of moderate/severe nonproliferative diabetic retinopathy (NPDR). Using intravitreal aflibercept-treatment (IVA) data from PANORAMA and DRCR Protocol W, the authors calculated the costs required to prevent proliferative diabetic retinopathy (PDR) and center-involving diabetic macular edema (CI-DME), and the costs required to improve the diabetic retinopathy severity score (DRSS) based on 2020 Medicare reimbursement data practice settings of hospital-based facility and nonfacility. The authors calculated these values to assess the cost-utility for NPDR treatment as these studies focused on the prevention and/or reduction of disease severity, rather than measuring visual acuity outcomes. This was a different measurement as previous comparative cost-effectiveness analyses were based on an assigned utility value that was derived from visual acuity changes that were then subsequently converted to quality-adjusted life-years. Notably, costs in this study were modeled based on 2020 Medicare professional and facility fee data. The study cost model assumed one initial new comprehensive eye code with remaining visits billed as intermediate established eye codes. The model assumed an OCT was performed at each visit and modeled injection costs with a J code. The authors calculated costs based on the following pertinent study findings from Protocol W and PANORAMA:

DRCR Protocol W Study: 399 patients with moderate to severe NPDR (DRSS 43-53) without CI-DME who were randomized to IVA or sham. Rate of progression to PDR at 2 years was reduced from 33.2% in the sham group to 13.5% in the IVA group. Rate of CI-DME with vision loss was reduced from 14.8% in the sham group to 4.1% in the IVA group. Improvement of 2 or more DRSS steps in 44.7% of IVA treated eyes compared to 13.7% of sham eyes.

PANORAMA Study: 402 patients with moderate to severe NPDR (DRSS 47-53) without CI-DME who were randomized to sham, 2 mg IVA every 16 weeks (2Q16), and 2 mg IVA every 8 weeks as needed (2Q8prn). 100-week results were analyzed. Compared to the sham group, the reported rate reduction of PDR was 9.1% in the 2Q16 group and 6.9% in the 2Q8prn group. For CI-DME, rate reduction was from 38.4% in the sham group to 11.3% in the 2Q16 group and 14.4% in the 2Q8prn group. The percentage of patients with an improvement of 2 or more DRSS steps was 13% in sham, 62% in 2Q16 group, and 50% in the 2Q8prn group.

The authors found that over 2 years in Protocol W, the cost required to prevent one case of PDR was $83,000 ($72,400) in the hospital facility (nonfacility) setting. In PANORAMA, the corresponding 2-year costs were $89,400 ($75,000) for the 2Q16 group, and $91,200 ($89,900) for the 2Q8prn group.

In Protocol W, the cost to prevent one case of CI-DME with vision loss was $154,000 ($133,000). In PANORAMA, the cost to prevent one case of CI-DME with or without vision loss was $70,900 ($59,500) for the 2Q16 group, and $90,000 ($88,800) for the 2Q8prn group.

The overall accumulated total for cost/DRSS unit change at the 2-year point was only calculated in Protocol W, as data for DRSS in PANORAMA was not available. This was found to be $2,700 ($2,400)/DRSS. This was $2,100 ($1,800)/DRSS in the first year and $6,100 ($5,300)/DRSS in the second year.

The authors conclude that there is a considerable cost associated with the prevention of PDR and CI-DME with IVA treatment of moderate/severe NPDR over 2 years, but also acquiesce that one cannot make assumptions about costs after this study period without further investigation. They suggest using the price per unit change in DRSS as a new parameter in future cost-utilization analyses.

This study (as with all cost-utilization analyses) brings up important issues to consider. The findings raise questions about whether value gained is worth the cost, and how the cost of preventing a particular complication might compare with the cost of treating a complication once it occurs (eg, PDR), both in the short-term and in the long-term. Further analyses and discussions are needed to develop metrics and benchmarks in the discussion of these kinds of treatments in the treatment of diabetic ocular disease.