4D-150

The ClinicalTrials.gov record for 4FRONT-1 (protocol 4D-150-C003) lists 95 sites: 91 in the United States and 4 in Canada. The study started on 2025-03-03; enrollment was completed in February 2026 and randomization in March 2026. Primary completion is estimated for June 2027 and study completion for June 2028, and the sponsor expects 52-week topline data in the second quarter of 2027.

According to the sponsor's trial schedule, participants in both arms receive aflibercept 2 mg at Weeks -5 and -1 and are randomized on Day 1, when the 4D-150 arm receives a single intravitreal injection of 4D-150 3E10 vg/eye and the aflibercept arm receives a sham injection. Both arms receive aflibercept at Week 4 (the third loading dose). From Week 12 through Week 52 the aflibercept arm receives aflibercept every 8 weeks and the 4D-150 arm receives sham injections at the same visits. Both arms receive a Durezol (difluprednate) taper starting on Day -3, and both arms are eligible for supplemental aflibercept when protocol-defined disease activity criteria are met.

No results have been reported as of September 2026; 523 participants were randomized and 52-week topline data are expected in the second quarter of 2027. The trial tests noninferiority rather than superiority for BCVA, with a margin of 4.5 letters for the FDA and approximately 4 letters for EMA and PMDA; a 4.5-letter margin is wider than the 4-letter margin prespecified in the faricimab (TENAYA and LUCERNE) and aflibercept 8 mg (PULSAR) registration trials. Participants in both arms can receive supplemental aflibercept, so the 52-week BCVA comparison measures 4D-150 plus as-needed aflibercept against fixed aflibercept every 8 weeks, and the key secondary endpoint (number of aflibercept injections) depends on the supplemental criteria, which exclude investigator discretion. The sponsor's 2024 design stated that no supplemental injections would be allowed in the control arm, whereas 2025 and 2026 disclosures state that both arms are eligible. Enrollment is restricted to treatment-naive participants in the United States and Canada with BCVA of 25 to 78 letters, CST of 500 µm or less, and a documented anatomic response to aflibercept during a run-in (15% or greater CST reduction or fluid resolution), which the sponsor describes as selecting strong aflibercept responders to reduce variability; results may therefore not apply to aflibercept non-responders, participants with thicker retinas, or previously treated populations. The primary endpoint is at 52 weeks, while 4D-150 is a one-time gene therapy intended for multi-year effect; durability beyond 104 weeks and long-term safety are not addressed by the registered outcomes. The statistical analysis sets, handling of intercurrent events, and multiplicity strategy have not been published.

The ClinicalTrials.gov record for 4FRONT-2 (protocol 4D-150-C004) lists 94 sites in 14 countries: United States (54), Argentina (6), Hungary (6), Japan (5), Australia (4), Germany (3), Latvia (3), Spain (3), United Kingdom (3), Lithuania (2), Singapore (2), Bulgaria (1), Italy (1), and Portugal (1). The EU trial record lists 9 member states (Bulgaria, France, Germany, Hungary, Italy, Latvia, Lithuania, Portugal, and Spain). The study started on 2025-07-22, enrollment was completed in June 2026, and 52-week topline data are expected in the second half of 2027.

According to the sponsor's trial schedule, which is shared by 4FRONT-1 and 4FRONT-2, participants in both arms receive aflibercept 2 mg at Weeks -5 and -1 and are randomized on Day 1, when the 4D-150 arm receives a single intravitreal injection of 4D-150 3E10 vg/eye and the aflibercept arm receives a sham injection. Both arms receive aflibercept at Week 4. From Week 12 through Week 52 the aflibercept arm receives aflibercept every 8 weeks and the 4D-150 arm receives sham injections at the same visits. Both arms receive a Durezol (difluprednate) taper starting on Day -3 and are eligible for supplemental aflibercept when protocol-defined disease activity criteria are met.

No results have been reported as of September 2026; enrollment was completed in June 2026 with more than 500 participants expected to be randomized, final randomization was expected in the third quarter of 2026, and 52-week topline data are expected in the second half of 2027 (registry primary completion estimated November 2028). The design matches 4FRONT-1 apart from population and geography, so the limitations of 4FRONT-1 apply: noninferiority testing with a 4.5-letter FDA margin and an approximately 4-letter margin for EMA and PMDA, supplemental aflibercept allowed in both arms under criteria that exclude investigator discretion, randomization restricted to participants with a documented anatomic response to aflibercept during the run-in, and a 52-week primary endpoint for a one-time gene therapy. Up to 40% of participants may be previously treated (1 to 4 prior anti-VEGF injections and diagnosis within 6 months), so results will not apply to patients with longer disease duration or extensive prior treatment. Because 4FRONT-2 enrolled across 14 countries, including sites in Europe, Latin America, Japan, Singapore, and Australia, regional differences in practice may add variability, and the EU trial record excludes women of child-bearing potential. The statistical analysis sets, handling of intercurrent events, and multiplicity strategy have not been published.

At the 24-week landmark (data cutoff January 19, 2024), the 3E10 vg/eye arm (n=20) had an 89% reduction in annualized anti-VEGF injection rate, 84% of participants received 0 or 1 supplemental injection, and 63% were injection-free; corresponding values in the 1E10 vg/eye arm (n=21) were 85%, 90%, and 50%. The average week 20 and 24 adjusted mean BCVA difference versus aflibercept 2 mg every 8 weeks (n=10) was -1.8 letters (95% CI -7.6 to 4.1) for 3E10 vg/eye and +1.8 letters (95% CI -3.8 to 7.4) for 1E10 vg/eye; confidence intervals include zero, and the study was powered for injection differences, not visual acuity. In the pooled Phase 1/2a 3E10 vg/eye population (n=24), treatment burden reduction versus the prior 12 months was 83% through year 1 and 79% through year 2 (data cutoff August 22, 2025), with a mean of 1.2 supplemental injections per participant in the 18 to 24 month interval.

The dose expansion is a small randomized Phase 2 study of 51 participants (20, 21, and 10 per arm) in which participants and investigators were not masked to 4D-150 versus aflibercept assignment and only outcome assessors were masked; it was powered for differences in anti-VEGF injections, not for visual acuity, and the BCVA and CST differences versus aflibercept have confidence intervals that include zero (for example, BCVA -1.8 letters, 95% CI -7.6 to 4.1, at 3E10 vg/eye). Comparative results versus the aflibercept arm have been reported only through week 24; later results pool the 3E10 vg/eye Phase 1 and Phase 2a participants (n=24) as a single-arm cohort, with treatment burden reduction calculated against each participant's injections in the prior 12 months rather than against a concurrent control. Two control arm participants received supplemental injections. The population had severe disease and a high prior treatment burden (mean CST 442 µm; 9.6 injections in the prior 12 months), which differs from the treatment-naive or recently diagnosed population of the 4FRONT Phase 3 trials. The Phase 3-eligible subgroup BCVA difference was reported as +3.3 letters in the company announcement, whereas the congress slides for the same subgroup show +2.4 letters. Results are available only from company announcements and congress presentations; no peer-reviewed full publication was identified, and no patient-reported outcomes have been reported.

At the 2-year analysis (data cutoff May 18, 2026), participants receiving 3E10 vg/eye (n=30) received a mean of 2.7 supplemental injections, a 78% reduction versus 12.0 injections projected for on-label aflibercept 2 mg every 8 weeks, and 46% were injection-free; the 1E10 vg/eye arm (n=15) received a mean of 3.7 injections (69% reduction). In the recently diagnosed subgroup at 3E10 vg/eye (n=15), the mean was 1.6 injections (87% reduction) and 66% were injection-free. The reduction at 3E10 vg/eye declined from 83% at 52 weeks (0.97 mean injections, 57% injection-free) and 82% at 1.5 years to 78% at 2 years. Mean BCVA change at week 52 was +2.2 letters with censoring of confounded assessments and +1.0 letter without censoring; mean CST change was -11 µm. At 2 years, BCVA and CST were described as maintained, with numerical values shown only in figures.

The Population Extension is an open-label, non-randomized cohort of 45 participants (30 at 3E10 vg/eye and 15 at 1E10 vg/eye) with sequential dose assignment and no concurrent aflibercept control; 13 of the 30 participants at 3E10 vg/eye came from an alternative steroid cohort, and three different prophylactic corticosteroid regimens were used. Treatment burden reduction is calculated against a projected on-label aflibercept every-8-weeks schedule (6.0 injections at 52 weeks and 12.0 at 2 years) derived from a single-arm model, not against observed injections in a comparator arm, and the 4D-150 arms received scheduled aflibercept at Day -7 and Day 28. Reported values changed across data cutoffs for the same timepoint (for example, 70% injection-free through 52 weeks by Kaplan-Meier estimate in September 2024 versus 57% at the January 2025 cutoff), and the 3E10 vg/eye reduction declined from 83% at 1 year to 78% at 2 years, with 46% injection-free at 2 years. The week 52 BCVA gain was +2.2 letters with censoring of cataract or surgery-confounded assessments and +1.0 letter without censoring; 2-year BCVA and CST results were presented only as figures. The recently diagnosed subgroup most comparable to the Phase 3 population includes only 15 participants. Aqueous humor aflibercept was below the limit of quantitation at any visit in 14 of 43 evaluable participants (33%). Pooled safety across 71 participants at 3E10 vg/eye is too small to detect uncommon events, adverse event frequency tables have not been released, and no peer-reviewed full publication or patient-reported outcome data were identified.

At 24 weeks (data cutoff April 3, 2023), 4 of 5 participants (80%) at 3E10 vg/eye and 2 of 5 (40%) at each of 1E10 and 6E9 vg/eye were anti-VEGF injection-free, with changes in annualized injection rate of -84%, -68%, and -80%. At 36 weeks (data cutoff July 3, 2023), all 4 evaluable participants at 3E10 vg/eye were injection-free and the lower doses had a 71% reduction in mean annualized supplemental injection rate. The 3 participants at 3E10 vg/eye who were injection-free beyond 52 weeks remained injection-free through approximately 2 to 2.5 years (data cutoff June 24, 2024), with mean BCVA +1 letter and mean CST -110 µm from baseline in that dose cohort. Aflibercept was detected in the aqueous humor of 10 of 11 samples (91%) at week 12.

The dose exploration stage is an open-label, uncontrolled study of 15 participants (5 per dose cohort), so dose-response and efficacy estimates rest on very small numbers; at the 36-week analysis, only 4 participants at 3E10 vg/eye were evaluable for supplemental injections, and BCVA analyses excluded participants with progressive cataract. Treatment burden reduction was measured against each participant's injection history in the prior 12 months, which is subject to regression to the mean in a population selected for 6 or more injections, rather than against a concurrent comparator. Baseline characteristics differed across cohorts (for example, mean BCVA 55, 65, and 75 letters and mean time since diagnosis 3.5, 2.6, and 4.7 years), and the population had severe, recalcitrant disease unlike the recently diagnosed or treatment-naive population of the Phase 3 trials. Long-term results after mid-2024 are reported only pooled with the Phase 2a dose expansion. Results are available only from company announcements and congress presentations, with no peer-reviewed full publication and no patient-reported outcomes identified.