The Event. Cue reported Q2 2026 and filed the 10-Q this morning. Headline: a $153.1M net loss, versus $8.5M in Q2 2025. Operating expenses of $95.6M on a company that burned $10.9M in Q1. On a screen, that looks like a catastrophe.
The Reality. Roughly $110M of that is non-cash or one-time: a $90.0M loss on issuance of liability-classified warrants (an artifact of the stock ripping from $14.74 on the April 30 pricing date to $29.85 at the May 4 close), $20.1M of fair-value recognition on the Ascendant pre-funded warrants and top-up obligation, and $19.9M of stock-comp on the Lin RSU grant I wrote about in July. Cash actually out the door for the license was ~$28M, of which $15.0M was the upfront.
The Signal. Buried in the same filing: the going-concern language I’ve flagged in every post since the deep dive is gone. Management now states cash on hand plus the July PIPE proceeds are “sufficient to meet its projected operating needs at least through the next twelve months.” That’s the auditors’ flag lifting — not a press release, just a sentence in Note 1.
The Tape. The stock closed $34.20, up 27.3% from $26.87 on August 3. It is now trading above the $33.21 Cormorant mark for the first time since the PIPE priced.
The $153.1M net loss decomposes into almost nothing that matters. The single largest line is $90.0M “loss on issuance of liability-classified warrants” — pure accounting mechanics. When Cue priced the May PIPE at $11.00 on April 30 and the stock closed at $29.85 by the May 4 closing, GAAP forces you to book the excess of instrument fair value over proceeds received as a loss. Translation: the company got punished on the income statement for its own stock going up 102% in four days. It was partially offset by a $24.5M gain as those instruments were remeasured down before stockholder approval reclassified them to equity on June 1. Net non-cash hit from the Ascendant transaction: $63.1M. The corresponding credit went straight to paid-in capital — no net impact on stockholders’ equity.
The going-concern determination is effectively remediated. On August 3 I wrote that “the auditors’ flag doesn’t clear until the next audited opinion” and that “anyone modeling this as a clean balance sheet is early.” The 10-Q moved faster than I expected: management now asserts twelve-month sufficiency without the substantial-doubt qualifier that appeared in the May 10-Q. $17.4M cash at June 30 plus $49.8M net from the July private placement gives roughly $67M of hard cash. Why it matters: the “structurally insolvent within six months” scenario from my deep dive now requires a CSU miss plus a failed follow-on, not just a CSU miss.
An amendment to the Ascendant license, signed today. Item 5 of the 10-Q discloses an August 14 amendment that “modifies the territories in which the Company has a non-exclusive right to manufacture CUE-221 to exclude China.” The implication: Cue’s ability to manufacture drug substance inside mainland China is being removed from the deal. Read the risk factor language right next to it — BIOSECURE-flavored exposure, “no assurance that the clinical data... will be accepted or considered by the FDA.” Narrowing the China manufacturing footprint appears to be exactly what you’d do to make a US IND and a future BLA cleaner, and to de-risk a Western acquirer’s diligence. It costs Cue optionality on cheap capacity. On balance this looks constructive for the regulatory path, and it is the second concrete step — after the food-allergy IND — toward answering the sharpest structural bear point I raised in the deep dive.
Driehaus Capital Management filed a 13G for 292,144 shares — 6.96%. Driehaus Life Sciences Master Fund is the holder of record for the bulk of it. That’s a dedicated healthcare fund crossing 5% as of June 30, joining Cormorant and Columbia Threadneedle on the register. Follow the money: three specialist or long-only institutions have now built positions in a name whose entire float was retail and warrant paper four months ago.
Both catalysts reaffirmed, again. CSU Phase 2 topline “by the end of the third quarter of 2026” — the fourth consecutive disclosure holding that line. CUE-401 IND and Phase 1 initiation “by year-end 2026.” Shares outstanding: 7,264,414 as of August 12, up from 4.05M in May. Add 2.59M common warrants and 1.56M options outstanding and the fully diluted picture is roughly 11–12M shares.
Nothing in this filing touched the biology, and that’s the point. CUE-221 is still the CD23-engaging anti-IgE — free-IgE neutralization at picomolar potency plus engagement of the CD23 feedback loop that tells B cells to stop making new IgE. Xolair does the first half only. The Phase 1 receipts are still n=3 per cohort across 4 dose levels (0.2, 0.6, 2, 10 mg/kg IV), with free-IgE suppression sustained beyond 12 weeks from a single dose.
The bar in ~6 weeks hasn’t moved. Omalizumab’s Phase 3 CSU data — UAS7 ≤6 at Week 12 in 51.9% (ASTERIA I), 65.8% (ASTERIA II), 52.4% (GLACIAL), and complete response (UAS7 = 0) in 33.7–44.3% — is a genuinely high bar for an active-comparator arm. Non-inferiority does not justify this tape. Incremental complete-response separation does. And ligelizumab remains the cautionary ghost: higher affinity, beat omalizumab on UAS7 in Phase 2b, then failed to differentiate in Phase 3. Anti-IgE is a class where Phase 2 superiority has a documented habit of not surviving a properly powered Phase 3, partly because CSU has a large placebo/regression-to-the-mean component and partly because the incumbent already works well in the majority of patients.
What the China amendment does and doesn’t do. It touches manufacturing rights, not the trial. The Phase 2 CSU study is still being run by Genesis Life Sciences — an Ascendant affiliate — in China, on a protocol Cue doesn’t control, producing data Cue doesn’t own. FDA acceptance of that dataset remains unresolved. What today’s amendment plausibly does is remove a China-based node from the eventual commercial supply chain, which matters if this asset is ever sold to a large-cap acquirer running BIOSECURE-aware diligence. De-risking the plumbing, not the data.
Three overlapping mechanics, and none of them are clinical:
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.