Continuous Ranibizumab via Port Delivery System vs Monthly Ranibizumab for Treatment of Diabetic Macular Edema: The Pagoda Randomized Clinical Trial

Abstract

Importance: Frequent visits and intravitreal anti-vascular endothelial growth factor (VEGF) injections are often required to manage diabetic macular edema (DME), burdening patients and their health care networks. The Port Delivery System (PDS) with ranibizumab is the first continuous anti-VEGF therapy that has the potential to reduce visit and treatment burden without sacrificing vision outcomes for patients with DME.

Objective: To evaluate the efficacy and safety through 64 weeks of ranibizumab, 100 mg/mL, via PDS with refill exchanges every 24 weeks (PDS Q24W) in patients with DME vs intravitreal injections of ranibizumab, 0.5 mg, every 4 weeks (monthly ranibizumab).

Design, setting, and participants: This randomized clinical trial was a phase 3, multicenter, noninferiority trial conducted across 87 sites in the US. Treatment-naïve patients at least 18 years old with center-involved DME were eligible for study participation. Enrollment was from September 30, 2019, to June 25, 2021; data were analyzed from September 30, 2019, to September 19, 2022.

Intervention: Participants were randomized 3:2 to receive 4 monthly doses of ranibizumab, 0.5 mg, followed by ranibizumab, 100 mg/mL, via PDS Q24W or monthly ranibizumab.

Main outcome and measure: The primary end point was change in best-corrected visual acuity (BCVA) from baseline averaged over weeks 60 and 64.

Results: A total of 634 participants were randomized (PDS Q24W group, n = 381; monthly ranibizumab, n = 253). The mean (SD) age at baseline was 60.7 (9.6) years; 363 (57.3%) participants were male and 271 (42.7%) female. Adjusted mean BCVA change from baseline averaged over weeks 60 and 64 was an increase of 9.6 letters for PDS Q24W and 9.4 letters for monthly ranibizumab (difference, 0.2; 95% CI, -1.2 to 1.6), meeting the primary end point of PDS noninferiority (margin, -4.5 letters). PDS Q24W participants had a mean (SD) decrease of 6.7 (12.0) letters 4 weeks after PDS insertion; mean BCVA was similar to monthly ranibizumab 16 weeks after implantation. Adverse events of special interest were more common in the PDS Q24W group (88 participants; 27.5%) than the monthly ranibizumab group (28 participants; 8.9%). No cases of endophthalmitis or retinal detachment were reported with PDS Q24W.

Conclusions and relevance: This trial found that changes in BCVA from baseline averaged over weeks 60/64 in the PDS Q24W group were comparable to the monthly ranibizumab group. While AESIs were more common with PDS Q24W, there were no instances of endophthalmitis or retinal detachment. Continuous ranibizumab, 100 mg/mL, via PDS was approved in the US for patients with DME in February 2025 and provides effective, durable, and generally well-tolerated treatment for DME with retreatment every 6 months through at least 64 weeks.

Trial registration: ClinicalTrials.gov Identifier: NCT04108156.

Figure 1.. Pagoda Randomized Clinical Trial Participant Allocation and Disposition (Efficacy Population)

PDS indicates Port Delivery System.

Figure 2.. Primary End Point and Vision Outcomes (Efficacy Population)

Difference in adjusted mean change in best-corrected visual acuity (BCVA) from baseline averaged over weeks 60 and 64 (A) and adjusted mean change from baseline in BCVA over time (B). 95% is a rounding of 95.05%; type I error was adjusted for interim safety monitoring. Adjusted means were estimated using a mixed-effects model. The model was adjusted for treatment group, visit, treatment × visit interaction, baseline BCVA Early Treatment Diabetic Retinopathy Study (ETDRS) letter score (continuous), baseline BCVA ETDRS letter score (<64 vs ≥64), previous intravitreal anti–vascular endothelial growth factor treatment (yes vs no), and diabetic retinopathy severity (nonproliferative vs proliferative diabetic retinopathy). An unstructured covariance structure was used. The estimate of the difference between the 2 groups used a composite contrast over weeks 60 and 64. The protocol-specified noninferiority lower bound margin was 4.5 ETDRS letters. Vertical bars represent 95% CI. PDS indicates Port Delivery System.

Figure 3.. Central Subfield Thickness (CST) and Diabetic Retinopathy Severity Score (DRSS) Outcomes Over Time

Adjusted mean change from baseline in CST (A) and proportion of participants with an Early Treatment Diabetic Retinopathy Study (ETDRS)–DRSS improvement of 2 steps or more from baseline (B). Vertical bars represent 95% CI. 95% is a rounding of 95.05%; type I error was adjusted for interim safety monitoring. CST is defined as the mean thickness of the central 1-mm circle of the ETDRS grid centered on the fovea. Adjusted means were estimated using a mixed-effects model. The model was adjusted for treatment group, visit, treatment × visit interaction, baseline best-corrected visual acuity (BCVA) ETDRS letter score (continuous), baseline BCVA ETDRS letter score (<64 vs ≥64), previous intravitreal anti–vascular endothelial growth factor (VEGF) treatment (yes vs no), and diabetic retinopathy (DR) severity (nonproliferative vs proliferative diabetic retinopathy [NPDR vs PDR]). An unstructured covariance structure was used. For DRSS, the weighted estimate was based on the Cochran-Mantel-Haenszel (CMH) method, stratified by baseline BCVA (<64 vs ≥64), previous intravitreal anti-VEGF treatment for DR with or without diabetic macular edema (yes vs no), and DR severity (NPDR vs PDR). Estimates below 0% or above 100% are shown as 0% or 100%, respectively. Baseline was defined as the last available value on or before randomization. Missing ETDRS-DRSS levels were imputed using the last observed ETDRS-DRSS level carried forward. PDS indicates Port Delivery System; mITT, modified intent-to-treat.