Confirmatory Evidence for All? FDA’s Approval of Lytenava for Wet AMD Shows How Non-Rare Diseases Can Use Confirmatory Evidence and That “Failed” Studies Can Provide Confirmatory Evidence

On July 22, 2026, FDA approved Outlook Therapeutics, Inc.’s (“Outlook”) Lytenava (bevacizumab-vikg) for the treatment of neovascular (wet) age-related macular degeneration (“Wet AMD”).  This approval was achieved after Outlook’s success through the Formal Dispute Resolution Request (“FDRR”) process following the receipt of multiple Complete Response Letters.  HPM’s Frank Sasinowski and Mark Tobolowsky are honored to have aided Outlook Therapeutics in this successful FDRR win and approval, and at the FDRR meeting, HPM’s Frank Sasinowski was the lead speaker representing Outlook’s positions.

Now that FDA has made the review documents public, the “Appeal Granted” memo, drafted by Dr. Mary Thanh Hai, Director of the Office of New Drugs, contains many useful lessons for sponsors.

Outlook submitted two pivotal studies as support for its BLA, which it referred to as NORSE 2 and NORSE 8.  NORSE 2 met its pre-specified primary endpoint; NORSE 8 did not.  This resulted in the December 2025 CRL that was subsequently appealed.  However, this situation was more than meets the eye.

In NORSE 2, subjects were randomized either to Lytenava dosed monthly or ranibizumab dosed once monthly for three months, and then every three months.  The primary endpoint assessed the proportion of subjects gaining ≥ 15 letters in BCVA from baseline to Month 11.  As described in Lytenava’s labeling, “the difference between the groups was significant, in favor of Lytenava.”  The Appeal Granted memo states, “[t]here is no disagreement between FDA and the applicant that NORSE 2 was an adequate and well-controlled clinical investigation where a statistically significant effect of bevacizumab-vikg on a clinically meaningful endpoint was demonstrated.”

As further stated in the memo, “[w]ith evidence of effectiveness from a highly persuasive adequate and well-controlled clinical investigation, at issue now is whether the applicant must conduct a second adequate and well-controlled clinical investigation or whether SEE [substantial evidence of effectiveness] can be established with the results from NORSE 2 plus confirmatory evidence.”  Ultimately, Dr. Thanh Hai concluded the latter.

Confirmatory Evidence from a Failed Trial

One of the key sources of confirmatory evidence supporting this approval was NORSE 8, which failed on its primary endpoint.  A failed study is not a typical source of confirmatory evidence, although not entirely unheard of.  However, as we learned in this appeal memo, it is crucially important that we do not wholly dismiss a study simply because it “failed.”  We must understand why a study “failed” and then determine whether and how it can still provide evidence of effectiveness despite the failure.

NORSE 8 was designed quite differently than NORSE 2 and any other Wet AMD efficacy study.  NORSE 8 was a much shorter study, designed to assess non-inferiority of Lytenava, as compared to ranibizumab, at Week 8 on the primary endpoint of mean change in BCVA.  For this study, the non-inferiority margin was defined to be -3.5 letters.  Both products followed the same dosing regimen through Week 12.

NORSE 8 failed to meet this pre-specified margin; at Week 8, the Lytenava group showed a mean increase in BCVA of 3.3 letters vs. a mean increase of 4.5 letters in the ranibizumab group.  The least squared mean difference (95% confidence interval) was -2.3 letters (-4.0, -0.5).  As the lower bound of the 95% confidence interval was -4.0, it exceeded the non-inferiority margin by 0.5 letters.

However, instead of dismissing the study entirely, the Appeal Granted memo took a closer look.  It was not clear where this -3.5 letter margin came from; it did not appear to be derived from studies comparing ranibizumab to sham, as might be expected.  Instead, it appeared to be derived from biosimilar study designs.  The purpose of a biosimilarity study is to test whether there are no clinically meaningful differences between two products by examining whether the two products demonstrate highly similar results at the most challenging timepoint to accomplish this.  As the Appeal Granted memo states, “[b]evacizumab-vikg, submitted under section 351(a) of the PHS Act, is not a biosimilar.”  In this context, the memo states, a less stringent non-inferiority margin might have been acceptable.  Nevertheless, the study cannot be declared a success by retrospectively applying a different margin.

The Appeal Granted memo stated that “interpretation of NORSE 8 must consider that the efficacy of bevacizumab-vikg was already demonstrated in NORSE 2 . . . after 11 months of treatment.”  As such, “[n]arrowly missing the prespecified NI margin after 2 months of dosing in NORSE 8 is not a compelling reason to discount the statistically persuasive longer-term efficacy findings in NORSE 2.”

Additionally, despite the missed primary, the visual acuity of Lytenava participants improved from baseline at every measurable timepoint and never differed from ranibizumab by a clinically meaningful difference (always within one line on the ETDRS chart).  Although the Week 8 results narrowly missed the non-inferiority margin, the Week 12 results converged and were within the margin.  As the memo concludes on this topic:

NORSE 8 failed to meet its primary objective to demonstrate noninferiority between bevacizumab-vikg and ranibizumab at Week 8, as specified in an agreed-upon protocol.  However, the results from this study are still interpretable and provide confirmatory evidence with regards to showing an improvement in BCVA with treatment where the difference favoring ranibizumab is not clinically meaningful, particularly had the analysis been performed at Week 12 after all three-monthly treatments were administered.  Finally, the 12-week study duration of NORSE 8 does not discount the longer-term efficacy findings from the 12-month study duration of NORSE 2 where efficacy was established at Month 11. (emphasis added).

The Appeal Granted memo states that a negative study should not generally be relied upon for confirmatory evidence.  However, the specific circumstances of this application, including that the data are still interpretable, rendered this appropriate to serve as confirmatory evidence.

Confirmatory Evidence from Natural History

The Appeal Granted memo also cited natural history evidence as providing another source of confirmatory evidence, which is a more typical source than a failed trial.  Here, there is strong natural history evidence to show that Wet AMD is a disease characterized by invariably progressive vision loss.  This well-understood natural history served to provide confirmatory evidence that the results in NORSE 2 are “real and clinically meaningful.”  As such, Dr. Thanh Hai concluded that “the natural history of [Wet AMD is] confirmatory evidence substantiating the treatment effect of bevacizumab-vikg observed in NORSE 2.”

Confirmatory Evidence from Mechanistic and Pharmacodynamic Data

Central foveal thickness (“CFT”), as assessed by optical coherence tomography (“OCT”), is a well-understood pharmacodynamic measure in Wet AMD.  CFT increases as a result of fluid leakage from macular neovascularization, and clinicians use OCT to measure CFT to assess response to therapy.  The mechanism of action of bevacizumab-vikg in treating Wet AMD is also well-understood.  As an anti-VEGF product, it is designed to prevent hemorrhaging and leakage caused by neovascularization.  This anti-VEGF activity was well-characterized by Outlook in functional and potency assays.

CFT was assessed in both NORSE 2 and NORSE 8 and showed deep reductions through the latest timepoint measured in each study.  The Appeal Granted memo clearly states that “CFT as measured by OCT is considered an accepted pharmacodynamic measure for evaluating the effect of VEGF-inhibition on fluid accumulation in the retina resulting from abnormal vessel leakage and hemorrhage and can serve as confirmatory evidence in this case.”  This application thus provided both mechanistic confirmatory evidence (anti-VEGF activity) and pharmacodynamic confirmatory evidence from NORSE 2 and NORSE 8 (reduced CFT), which both provide confirmatory evidence to support NORSE 2.

Parting Thoughts

The Lytenava FDRR win and subsequent approval is remarkable in a few ways and provides some important lessons for sponsors who might find themselves similarly situated.

One Adequate and Well-Controlled Study Plus Confirmatory Evidence Is Not Just for Rare Diseases

First, although the single adequate and well-controlled study and confirmatory evidence pathway has been most often applied by FDA to applications for rare disease indications, that was not the case here.  Wet AMD is not a rare disease.  The statutory language added by FDAMA 115 creating this pathway does not limit it to rare disease applications, but that has long been how it has applied.

FDA’s recent June 2026 update to the Substantial Evidence of Effectiveness Guidance has been framed as expanding this “One Adequate and Well-Controlled Clinical Investigation Plus Confirmatory Evidence” pathway to all products.  It is too early to tell if FDA’s approach will shift as a result of this guidance.  Notably, the Appeal Granted memo does not cite to this guidance, instead only referring to the September 2023 draft guidance, “Demonstrating Substantial Evidence of Effectiveness with One Adequate and Well-Controlled Clinical Investigation and Confirmatory Evidence.”  As the CRL being appealed in the FDRR pre-dated the June 2026 updated guidance, perhaps it was considered out of scope of the appeal, as FDRRs are intended to be based on the available information at the time of the relevant decision.

Nevertheless, this approval shows that the one study plus confirmatory evidence pathway is not just for rare diseases.  If the framing of the new June 2026 guidance is accurate, perhaps we will see more such approvals down the road.

A Failed Study Can Contribute to SEE

The second key takeaway here is that a failed study is not necessarily unable to contribute to substantial evidence of effectiveness.  A rigid regulatory approach that treats it as such in all cases is inappropriate.

As was demonstrated in this appeal, it is crucially important to understand what the study can contribute beyond a missed primary.  The best approach is not to retcon a failed study into a successful one had it been differently designed, but it is rather more nuanced than that by evaluating what the evidence actually demonstrates and telling this story in an appropriate and coherent way.

Here, the Appeal Granted memo appropriately assessed what the study was designed to show and the evidence it actually provided beyond the successful/failed rubric.  With this understanding, the study was still able to provide confirmatory evidence of the results of the positive study despite the failed primary endpoint.  While the use of a failed study as confirmatory evidence is not entirely unprecedented, it is exceedingly rare, and we were extremely pleased to see this wonderful outcome for patients.

Onward!