The 9 Patients Who Cost EyePoint $1 Billion — And What Remains

EyePoint's CEO told investors on August 5, 2026 that the upcoming LUGANO trial results "represent transformative advancement in the wet AMD landscape." Thirteen days later, he disclosed that the drug had failed its primary endpoint. The stock fell 67% in one session. On August 20, a securities law firm asked investors whether the company's confident language had misled them.

The number that matters is not the 67% drop. It is 9 out of 211.

EyePoint's lead drug candidate, DURAVYU, missed its vision endpoint by the thinnest margin a Phase 3 trial can produce: nine patients in the treatment arm lost significant vision from causes the company says were unrelated to wet AMD.No patients in the control arm experienced the same outcome. The company argues the control group outperformed what any aflibercept trial historically achieves — about 3% to 5% of control patients should lose that much vision, yet zero did in LUGANO. The statistical math follows: one arm performed identically to the drug; the other arm performed better than the drug because the control was unusually good.

That is the claim. The number is not proof it is correct, but it is the door.

How a non-inferiority trial can fail on a coin flip

To understand what happened, you need to know how a non-inferiority trial works. EyePoint was not trying to prove DURAVYU is better than aflibercept, the current standard injection treatment for wet age-related macular degeneration. It was trying to prove DURAVYU is not worse — while delivering the drug every six months instead of every one to two months. The trade-off is the entire business model: fewer injections for patients, dramatically lower treatment burden, a commercial product that could displace the dominant therapy.

The primary endpoint measured whether vision was preserved equally between the two arms over 56 weeks. The result: not achieved.

But nine patients drove the entire failure. Here is the breakdown EyePoint disclosed: six eyes lost vision to geographic atrophy, a degenerative condition. One each to glaucoma and detached retina. EyePoint's CEO said three of the nine had their underlying wet AMD well controlled — the vision loss came from something else entirely.

The control arm, meanwhile, produced a statistical anomaly. Across dozens of published Phase 3 trials, aflibercept patients lose significant vision at a rate of 3% to 5%. In LUGANO, the rate was zero. When the bar on one side of a non-inferiority test happens to stand taller than it normally does, the test becomes harder to pass even if the drug works.

EyePoint ran an ad hoc analysis — a post-hoc statistical exercise, not a pre-planned test — excluding those nine patients. In that stripped-down dataset, DURAVYU met non-inferiority with a nominal p-value of 0.0096. The FDA does not approve drugs on ad hoc analyses. That number is useful context, not a regulatory path.

The secondary endpoints tell a different story. DURAVYU cut treatment burden by 42% versus aflibercept, with a p-value below 0.0001. That translates to roughly two fewer injections per patient over 56 weeks. Fifty-four percent of DURAVYU patients needed zero supplemental injections through Week 56. Central retinal thickness — the anatomic measure of fluid control — was within 4 microns of aflibercept, essentially identical. No new safety signals emerged: no insert migration, no retinal vasculitis, no inflammation.

So you have a drug that appears to maintain retinal anatomy, dramatically reduces injection frequency, and shows a clean safety profile — but could not cross the primary vision threshold in one trial because of a handful of confounding patients and an unusually strong control arm.

The allegation is language, not accounting

This is not an accounting investigation. There are no suspicious cash flows, no revenue recognition issues, no related-party transactions to trace. EyePoint is a clinical-stage biotech with $0.5 million in quarterly revenue, $94.5 million in quarterly losses, and a balance sheet funded almost entirely by cash and marketable securities.

The securities claims focus on what management said before the results came out. On August 5, the CEO called the trial "transformative." In earlier quarters, he characterized DURAVYU as having "best-in-class potential" through a "clinically rigorous, de-risked" approach. The company told investors its Phase 3 program was "on track" in the days before the August 17 disclosure.

Forward-looking statements carry Safe Harbor protections under federal securities law. A CEO can say a drug looks promising without guaranteeing results. The question these investigations will probe is timing and specificity: did management know, or should it have known, that the primary endpoint was in jeopardy before telling investors everything was proceeding as planned? That is a question for regulators and courts, not for an earnings model.

At this stage, this is Level 1 on the evidence ladder: a discrepancy between optimistic language and a disappointing outcome. The company did not hide the failure — it disclosed it. The legal claims are in their earliest phase, filed by firms whose revenue model depends on finding shareholders who lost money and a defendant who can pay a settlement.

The real investor question is cash, not claims

The securities suits are noise. The signal is the balance sheet.

EyePoint had $180 million in cash and marketable securities as of June 30 — $110 million in cash and $70 million in marketable securities. Its Q2 net loss was $94.5 million, though operating cash outflows are moderated by non-operating investment income. The company said in its Q2 filing that its cash position funds operations "into the fourth quarter of 2027."

That runway is the constraint. At a current quarterly burn rate near $100 million in operating expenses, and with two more pivotal datasets to fund — the LUCIA trial for wet AMD (data expected Q4 2026) and the COMO/CAPRI trials for diabetic macular edema (data expected Q4 2027) — EyePoint is burning through its bridge to approval.

The company's market capitalization sits at roughly $410 million, down from approximately $1.5 billion before the crash. The stock, trading around $4.58, is down 75% year-to-date. Capital flows show persistent net outflows across every investor size: block, large, medium, and retail orders have all tilted to the sell side. The institution that bought a story of transformative advantage is now watching a story that needs one more trial to survive.

What the LUCIA trial decides

The entire case now hinges on a single dataset: the second pivotal wet AMD trial, LUCIA, with identical design to LUGANO. Topline data is expected in the fourth quarter of 2026.

If LUCIA confirms LUGANO — another primary miss with a similar asymmetric pattern — the FDA will view this as a consistent failure, not a one-trial statistical fluke. The NDA filing EyePoint planned for the first half of 2027 would face serious doubt, and the company would need to raise capital, restructure its pipeline, or pursue a business development deal.

If LUCIA succeeds — meeting the primary endpoint and confirming the secondary treatment-burden advantage — the LUGANO miss becomes a footnote. Two of two pivotal trials is what the FDA requires, and one anomalous trial paired with one clean trial has happened before in retinal drug development. EyePoint could proceed to NDA filing and the stock could partially recover.

If LUCIA is inconclusive — a borderline result that neither confirms nor denies the drug's efficacy — EyePoint faces the worst outcome: cash running out while the regulatory path remains ambiguous. In that scenario, the company would need to raise capital at a depressed valuation, diluting existing shareholders, or pursue a partnership that would give away margin for survival.

The shareholder invoice

Investors who bought at the peak — when the stock traded near $19 and the market valued the company at roughly $1.5 billion — have seen more than a billion dollars evaporate. The 67% single-day decline on August 17 erased approximately $940 million in market value in one afternoon.

But the shareholder cost is not just the paper loss. It is the time value of capital tied to a binary outcome that may not arrive for another six months. The $180 million in cash that funds the company's operations through late 2027 belongs to the shareholders, and that cash will be consumed regardless of whether DURAVYU succeeds or fails. If the company needs to raise more capital to survive a negative LUCIA result, that cash will be raised at prices set by the post-LUGANO market.

The current price of $4.58 against a pre-crash analyst target of $36.62 does not represent undervaluation. It represents the market's assessment that the probability of approval has moved sharply lower — from a near-certain story that justified a $1.5 billion valuation to a coin flip that may not land in the company's favor.

What to make of it

EyePoint is not hiding numbers. It is a development-stage company whose entire valuation rested on one drug's ability to clear one regulatory hurdle, and that drug stumbled. The company disclosed the failure promptly and explained its theory — that nine patients and a strong control arm produced a statistical miss, not a therapeutic one.

The securities investigations are standard post-crash activity. They will proceed or settle based on whether investigators can show management knew about the endpoint risk before the confident statements. That is an inquiry into timing and knowledge, not into the science.

The investment case reduces to a single question you cannot answer yet: will LUCIA confirm the drug works or confirm the miss? The fourth quarter of 2026 will tell. Until then, EyePoint is a company spending $100 million per quarter to find out whether it has a business — and the cash to survive the answer runs out in the same year the answer arrives.