Alejandro Perez Pitarch 1, Sebastien Bihorel 1, Kenneth Turner 1, Joachim Hoechel 2, Thomas Eissing 2, John Davis 1, Lutz Harnisch 1
1 Regeneron Pharmaceuticals (Tarry Town, USA), 2 Bayer AG (Leverkusen, Germany)
Introduction/Objectives: Aflibercept is a fully human recombinant fusion protein that acts as a soluble trap for vascular endothelial growth factor and inhibits its binding to and activation of endogenous receptors. Findings from the pivotal PULSAR and PHOTON trials supported the regulatory approval of aflibercept 8 mg for neovascular age-related macular degeneration (nAMD) and diabetic macular edema (DME) in several countries, including those of the European Union, Japan, and the United States, where it is also approved for diabetic retinopathy.
In the pivotal PULSAR and PHOTON studies for nAMD and DME, respectively, patients were randomized to aflibercept 8 mg every 12 or 16 weeks following 3 initial monthly doses. Per protocol, the minimum dosing interval permitted was Q8. An exploratory post hoc analysis found that 5.8% and 1.4% of patients in the PULSAR and PHOTON studies, respectively, would have met criteria for further dosing interval shortening to every 4 weeks (Q4) dosing at week 48 or 96, had it been allowed per protocol.
The primary objective of this analysis was to elucidate the characteristics of patients with DME or nAMD who may have benefited from aflibercept 8 mg Q4 in the PULSAR and PHOTON studies, and to infer the magnitude of this benefit.
Methods: The analysis was limited to patients treated with aflibercept 8 mg in the the PULSAR and PHOTON studies. Pharmacokinetic (PK)/pharmacodynamic (PD) models were developed to characterize the relationships between aflibercept concentrations in the eye, central retinal thickness (CRT) and best corrected visual acuity (BCVA) longitudinally. Aflibercept concentrations in the eye (governed by ocular clearance [QE]) were utilized in this analysis to integrate the previously developed population PK model with the response of CRT and BCVA over time.
Simulations were performed to predict the impact of dosing interval adaptations (in response to changes in CRT and BCVA) on treatment effect and clinical outcomes. In particular, the clinical benefit of shortening to Q4 in patients who still met the DRM criteria while on aflibercept 8 mg Q8 in the PULSAR and PHOTON studies was investigated. In addition, the clinical benefit of interval shortening to Q4 in virtual cohorts resembling theoretical future treated populations with nAMD or DME was investigated.
Results: The PK/PD models successfully described the observed data and characterized the relationship between predicted aflibercept concentrations in the eye, CRT, and BCVA over time. The models allowed for the investigation of the duration of effect of aflibercept 8 mg dosing across the range of QE values observed in the studies. The analysis presented herein indicates that a small proportion of patients, characterized by high baseline CRT and fast QE, are more likely to experience oscillations in CRT and BCVA than other patients with DME and nAMD. Most patients maintained effective drug concentrations for a sufficient period of time; hence, 8-week dosing intervals and longer were adequate. However, a small proportion of patients may require more frequent dosing (Q4) to maintain effective concentrations and prevent loss in efficacy. Simulations based on the patients on a Q8 dosing interval in PULSAR and PHOTON who met the DRM criteria for hypothetical shortening to Q4 suggest that these patients would have seen a benefit of 2.5 or 5.5 letters on BCVA with Q4 dosing, respectively.
Conclusions: This analysis found high baseline CRT and fast QE to be associated with the need for more frequent dosing (Q4). Model-based simulations inferred that patients on aflibercept 8 mg Q8 who still met the DRM criteria in PULSAR and PHOTON may have benefited from Q4 dosing.
Reference: PAGE 34 (2026) Abstr 12138 [www.page-meeting.org/?abstract=12138]
Poster: Drug/Disease Modelling - Other Topics