Faricimab 6 mg versus aflibercept 8 mg in diabetic macular oedema: an in silico study of anchored efficacy and common-protocol treatment burden

Background No prospective randomised head-to-head trial has compared faricimab 6 mg with aflibercept 8 mg for diabetic macular oedema (DMO). Their pivotal programmes used materially different dosing architectures, precluding interpretation of observed interval distributions as drug properties. Methods We separated an anchored Bayesian aggregate-data comparison of year-1 and year-2 best-corrected visual acuity (BCVA) and retinal-thickness differences from an individual-level, discrete-time burden microsimulation. The latter applied identical three-dose loading and Q4W–Q16W rules, conditional on transported durability evidence. Protocol standardisation applied only to burden; dosing history did not generate efficacy. Inputs represented mixed pivotal-trial populations. Structural, prior, composite leave-out, qualification and calibration analyses were performed. Results The estimated BCVA difference between treatments was small and compatible with no treatment difference: faricimab minus aflibercept 8 mg was +1.22 letters at year 1 (95% uncertainty interval [UI], −0.48 to +2.86) and +0.41 at year 2 (−1.43 to +2.27). Retinal-thickness-reduction differences were +21.93 μm at year 1 (+1.70 to +42.52; P[Δ > 0] = 98.3%) and +20.53 μm at year 2 (−1.08 to +42.24); the year-2 interval included zero, and estimates varied across structures. Under the common simulated treat-and-extend regimen, aflibercept 8 mg was projected to require fewer injections than faricimab (7.29 vs. 8.01 by week 52 and 10.93 vs. 12.60 by week 104). A greater proportion of simulated eyes receiving aflibercept 8 mg than faricimab reached 16-week dosing intervals (80.2% vs. 50.9% at week 52 and 77.7% vs. 60.4% at week 104). Aflibercept 8 mg showed a lower estimated injection burden, although the adjusted range included no difference between treatments. Conclusion The estimated BCVA differences were small and compatible with no treatment difference. Faricimab showed greater anatomical improvement at year 1, while the year-2 estimate included no difference. Aflibercept 8 mg showed a lower simulated injection burden, although the adjusted range included no difference. Neither treatment demonstrated clinical superiority.