NRx Pharmaceuticals ($NRXP): What Is the November 10 GDUFA Date Actually For?

01 Bull, base and bear scenarios

These are reading frames, not forecasts, and none of them is a recommendation. Each is written so that a reader can say which document would confirm or break it.

What distinguishes NRx from most small-cap biotech names is that the near catalyst is not a clinical readout and is not binary in the usual way. A generic application for a drug already on the market, whose only open objection concerns a container closure, is a manufacturing and paperwork question with a date on it. That makes the November 10 event more legible than a trial result — and it also means the balance sheet, not the science, is what decides whether the company gets to the other side of it on its own terms.

What Would Falsify This Reading

The reading on this page is that one date is set by the agency, though known through the company’s report, and the rest are intentions, that the balance sheet is tighter than the announcements suggest, and that the gap between what is said and what is filed is itself the thing to watch. Each of the following would damage that reading, and each is checkable against a document.

• The review history is better than the headline, not worse. The first cycle on the abbreviated application produced no major deficiency relating to the drug’s components — the objection is a twist-off cap on a luer lock vial. After a clarification meeting attended, the company says, by the Office of Generic Drugs and senior leadership of the drug centre, the agency reclassified the amendment from major to minor on September 15, 2026 and issued an acknowledgement letter with a November 10 goal date. A minor amendment carries a three-month review goal; a major one would have carried six to ten months. The FDA’s letters are not public, so this history is the company’s account of them. If that date holds, this stops being a story about announcements.
• There is revenue, and it did not exist a year ago. Net patient service revenue was $1.101 million in the June quarter and $2.169 million over the half, against zero in both 2025 comparatives, from six clinics in Florida. Cost of patient services was $675 thousand in the quarter. That is a small number, but it is a real one on a filed income statement, and it changes what the company is.
• The debt is genuinely gone. There are no notes, no convertibles and no bank borrowings. The only financial debt on the balance sheet is a $412 thousand insurance loan. The residual $5.4 million owed to its former lender was converted to equity in December 2025. The $18.550 million of warrant liabilities that makes the liability total look large is a fair-value derivative, not borrowed money, and the $5.8 million Alvogen repayment obligation sits in accruals.
• The branded application may not need new trials. At a Type C meeting on February 11, 2026 attended by a Deputy Director of the drug centre and the directors of the neuroscience office and psychiatry division, the agency directed the company toward full approval rather than accelerated approval and a broader indication, and agreed to review patient-level data from completed trials as substantial evidence of efficacy, with real-world evidence as confirmatory support; in the annual report’s words, the agency “did not ask the Company to conduct additional clinical trials.” This is the company’s account of a meeting whose minutes are not public, and the statistical plan for the real-world data was still being agreed in August. If it holds, the branded path is cheaper than a reader would assume.
• The DARPA item has moved. On September 30 the defence subsidiary announced “the award of an initial $11,177,757.27 contract” for the SPARC-TMS trial. If a Form 8-K follows with the contract and the registry shows enrolment, the defence trial stops being an intention and becomes funded work.
• Or the balance sheet decides it before the FDA does. This is the strongest argument against the rest and it comes from the company’s own filing. Substantial doubt exists about going concern. Liabilities of $34.330 million exceed assets; the stockholders’ deficit is $170 thousand and working capital is negative $4.7 million. Cash was $26.686 million against roughly $3.0 million a month of burn, the share count rose 36 per cent in under eight months, and the two runway sentences published with the same results do not agree. An approval reached after a raise at the current price is a different outcome for existing holders than an approval reached without one.

None of these is a prediction. They are the observations that would make the rest of this page wrong, listed so that a reader can check them rather than take the reading on trust.

02 Executive answer: one certified date, and the rest

Separating one certified date from a stack of stated intentions

This company announces a great deal, and the announcements are not all the same kind of thing. One of them is a goal date set by the FDA in an acknowledgement letter. One is a defence-agency selection that the July filing said was not an award, followed on September 30 by a press release announcing an initial contract that no filing yet records. One is a spin-off described in a risk factor the company may “pursue, modify or abandon… at any time.” This page sorts them, gives each its document and its date, and says plainly where the record is silent — including on the three things a reader would most expect to find and will not: a confirmed Nasdaq compliance notice, a Form 4 in the last six months, and any registration for the clinic business.

NRx Pharmaceuticals is a clinical-stage company that has recently started earning revenue from a clinic business, with two applications in front of the FDA and no approved product. Six facts set the frame.

There is one date set by an agency, and it is close. The abbreviated application for preservative-free intravenous ketamine — the generic path, branded KETAFREE — has a GDUFA goal date of November 10, 2026. It was set by an acknowledgement letter after the FDA, on September 15, 2026, reclassified the one remaining deficiency from major to minor. The first review cycle had found, by the company’s account, no major deficiency in the drug’s components; the single objection concerns the twist-off cap on a luer lock vial. The FDA does not publish its letters on generic applications, so the date and the history reach readers through the company’s Forms 8-K.

The DARPA item is an announced contract, not yet a filed one. On July 21, 2026 a subsidiary received a letter saying its proposal had been selected for negotiation of a potential award, and the Form 8-K stated that this “does not constitute a notice of award, a commitment by DARPA to make an award or an authorization for the Company to incur costs.” On September 30, 2026 the subsidiary announced “the award of an initial $11,177,757.27 contract,” close to the “more than $11.5 million” the company had guided. No Form 8-K had been filed by October 5, the contract terms are not public, and the trial is listed as not yet recruiting past its own estimated start date of September 1, 2026.

The company states that substantial doubt exists about going concern. At June 30, 2026 there was $26.686 million of cash and no marketable securities, against total liabilities of $34.330 million, a stockholders’ deficit of $170 thousand and a working-capital deficit of $4.7 million. The auditor added a going-concern paragraph to the 2025 annual accounts. Accumulated deficit is $324.845 million.

Two runway sentences were published three days apart and they do not match. The filing says the cash funds drug-development operations “through at least the first quarter of 2027”; the press release on the same results says “at least one year.” Both appear verbatim later on this page. Operating cash use was $13.170 million over the half, of which the June quarter accounted for $8.875 million — about $3.0 million a month.

The short position is unusually large for a company this size. 9,773,562 shares at the September 15 settlement, about 22.6 per cent of the 43,295,767 shares outstanding, at 16.27 days to cover. At the February 27 settlement it was 2,459,826. That is close to a fourfold increase in seven months, and the days-to-cover figure has tripled because the reported average daily volume in that series fell while the position grew.

There is now revenue, and the clinic business is the only part of the company that generates any. Net patient service revenue was $1.101 million in the June quarter and $2.169 million over the half, against zero in both 2025 comparatives, from six clinics in Florida held through a majority-owned subsidiary. A partial spin-off of that subsidiary has been announced since 2024; it has no registration statement, no date, and a risk factor reserving the right to abandon it.

03 The generic application, and the vial cap

This is the part of the file with a date on it, so it is worth following in sequence. The product is preservative-free intravenous ketamine in a single-dose presentation, pursued as a generic of an existing reference product rather than as a new drug.

What is not established

Three things a reader might assume are not in the record and are not asserted here. No approval, tentative approval or refusal has been issued — November 10 is a goal date, not a decision. No revenue figure for the product exists, because it has not been sold. And no filing confirms that the manufacturer certifications requested on August 6 were accepted as sufficient; the reclassification on September 15 is strong evidence that they were, but it is inference from an outcome rather than a statement of acceptance.

Two further points sit in the filings rather than the headlines. The company’s documents never use the words “complete response letter” for the July 30 outcome; under the GDUFA framework a review cycle that ends with an open major deficiency is normally closed with such a letter, answered by an amendment, which fits the “amendment” language of the September 15 release. That reading is a Merlintrader inference, not a disclosure. And the 10-Q reports that an FDA inspection of the manufacturing facility ended with a Voluntary Action Indicated classification, which the company presents as meeting the pre-approval requirement. In the FDA’s own definition, VAI means the inspection “found objectionable conditions or practices but the agency has determined the facility can voluntarily correct its deficiencies and will not recommend any action.” The annual report also records a manufacturing-readiness audit at Nephron Pharmaceuticals with no deficiencies, and the company has filed a citizen petition asking the FDA to remove the preservative benzethonium chloride from all ketamine presentations. FDA inspection classifications → GDUFA III goals →

04 The branded application, the defence trial, and the clinics

Everything in this section is either a company intention, a registry record, or an agency meeting outcome. None of it has a date that an agency has committed to, and one item is routinely described as something it is not.

NRX-100, the branded path

The same molecule, pursued as a new drug application rather than a generic, for suicidal ideation in depression. The application was started in the fourth quarter of 2024 and has never been completed. The annual report filed March 23, 2026 guided the filing to the second quarter of 2026; it did not happen. The August 17, 2026 release says the company is “poised to finalize its NDA, aiming for 2027 approval,” and the quarterly report filed three days earlier is more specific: it anticipates submitting the application by the end of the third quarter of 2026. That quarter closed on September 30 and no filing announces a submission, which makes it the third missed self-set date on this programme in under a year: the 10-Q of November 2025 had expected the application “to be completed in the fourth quarter 2025.” The company’s own explanation, in the 10-Q, is that “filing was delayed while the Company resolved the luer lock vial issue because the same vial is used in NRX-100,” which ties the branded timetable to the generic review. There is no PDUFA date, because there is no filed application, and there has been no complete response letter and no refuse-to-file in the record.

What did happen is a meeting worth understanding, because it changed the shape of the path. At a Type C meeting on February 11, 2026 attended by a Deputy Director of the drug centre and the directors of the neuroscience office and the psychiatry division, the agency, according to the company’s annual report and a February 17 release that cites “oral guidance received at the meeting”:

• directed the company toward full approval rather than accelerated approval;
• pointed it at a broader indication — patients with severe depression, including bipolar, who may have suicidal ideation, rather than only those with active ideation;
• agreed to review patient-level data from the completed trials as substantial evidence of efficacy, with real-world evidence as confirmatory support, and, in the annual report’s words, “did not ask the Company to conduct additional clinical trials.”

Agreement to review is not agreement that the evidence is sufficient, and the meeting minutes are not public. The 10-Q adds that the company and the FDA were still “in the process of reaching agreement on the statistical methods” for the real-world evidence, with a Type A meeting planned to settle them.

The real-world dataset is the unusual asset here. The company puts it at more than 65,000 patients treated with intravenous ketamine against about 6,000 with intranasal esketamine, with an interim analysis of the first 20,000 suggesting faster onset and larger effect. Separately, Fast Track designation was broadened on August 8, 2025 to suicidal ideation in depression including bipolar depression, from a 2017 designation limited to bipolar depression, and the letter contains an explicit unmet-medical-need finding — a qualifying condition for a national priority voucher programme the company says it has applied to. No outcome of that application exists in the record.

NRX-101, the oral combination

An oral D-cycloserine and lurasidone combination carrying Breakthrough Therapy designation from November 2018, Fast Track from 2017, and a Special Protocol Agreement from April 2018. Its application has been started with the manufacturing module only, with a request for rolling review. The reported Phase 2 result is a statistically significant reduction in depression at p=0.04 and in suicidal ideation at p=0.02 against lurasidone alone over 42 days. The licensing milestone remaining is modest: $187,500 on first FDA approval. Of the registered trials led by the NeuroRx subsidiary, three are completed, three carry an unknown status with no update since November 2023 or January 2024, and one, a 2021 vaccine study, was withdrawn; one expanded-access record remains available.

SPARC-TMS and the DARPA letter

HOPE Therapeutics, the clinics, and the spin-off that has no date

This is the only part of the company with revenue. HOPE was formed in February 2024 and NRx describes itself as founder and majority owner — a change from the “wholly owned” language used through 2025, which is itself worth noting.

Two further subsidiaries were formed in 2026, neither with revenue: a defence-systems entity, and a monoclonal-antibody entity. The second is not an empty shell — the filing says it carries one clinical-stage candidate with data from three completed human trials, in multiple sclerosis, type 1 diabetes and long covid, alongside a preclinical programme in motor neurone disease whose first human dose is targeted for July 2027. NRx bought these assets as the winning bidder in a Swiss liquidation sale of GeNeuro SA in May 2026, and the company says the subsidiaries are meant to be funded from sources that do not dilute NRx shareholders.

The investor presentation furnished on September 14, 2026 is the most detailed public plan for that subsidiary. It targets a Phase 1/2 start in July 2027, a Phase 3 start in January 2028 and an NDA submission in January 2029, and lists “potential financing” of a $5 million Regulation CF raise and $3 million from the Congressionally Directed Medical Research Programs in 2026, and a $75 million Regulation A raise in 2027 upon a human IND. It also cites $80 million added to the 2027 defence budget for ALS trials; the 10-Q is more careful, saying there is “no allocation specific to GeNeuro/NIH for these funds.” Manufacturing of human-grade antibody had not started at the time of the 10-Q. Form 8-K, Sept. 14 →

05 Cash, burn, and two different runway sentences

The burn, separated properly. Operating cash use was $13.170 million over the first half. The first quarter accounted for $4.295 million of that, so the June quarter consumed $8.875 million — about $3.0 million a month, calculated rather than disclosed. The quarterly net loss was $16.5 million and the half-year loss $17.9 million, both much larger than the operating loss of $6.505 million and $11.250 million because of non-cash warrant revaluation. Research and development was $2.116 million in the quarter; selling, general and administrative $4.745 million. Capital spending added $0.897 million over the half. For the full prior year, the net loss was $28.622 million and operating cash use $14.112 million.

Reading the deficit correctly. The stockholders’ deficit narrowed from $15.937 million at December 31, 2025 to $170 thousand at June 30, because $33.7 million added to paid-in capital, almost all of it from the at-the-market sales and the June offering, more than covered the $17.95 million loss. Of the $33.874 million of current liabilities, $18.550 million is the warrant liability, which is settled in shares if the warrants are exercised rather than in cash; strip it out and working capital is positive by about $13.8 million (a Merlintrader calculation). The other $15.324 million of current liabilities is $3.146 million of payables, $11.048 million of accrued and other liabilities including the $5.8 million owed to Alvogen, $0.340 million of accrued clinical-site costs, the $0.412 million insurance loan and $0.378 million of current lease liabilities.

06 The capital structure, and what the last twelve months have cost

That is a 36 per cent increase in under eight months, and almost all of it came from two places: the at-the-market programme (4,812,972 shares) and the June offering (6,378,942). The equity statement adds 218,161 shares issued for services in May, 68,161 of them to a licensor, plus 10,000 vested restricted shares and 3,336 from a cashless warrant exercise. Authorised capital is 500,000,000 shares; no preferred is outstanding, the 2023 issue having converted in March 2024.

Where the money came from

Warrants, options and the fully diluted count

One governance detail sits underneath all of this. The equity plan’s annual evergreen was raised at the March 2026 meeting from the lesser of 1 per cent of fully diluted shares or board discretion, to the lesser of 3,187,234 shares, 5 per cent of fully diluted, or board discretion. That proposal drew 4,484,064 votes against 5,976,632 in favour — 42.9 per cent of the votes cast were against it, which is a notably high dissent for a plan amendment and is the one item on the ballot where shareholders split.

The arithmetic of the plan reserve is worth a line. The 10-Q puts the shares still available under the 2021 plan at 320,660 at June 30, 2026, after a January evergreen top-up of 417,143. The July 28 awards described as made under the plan total 423,353 — 398,353 units and 25,000 options — which is about 102,700 more than the reported reserve (a Merlintrader calculation). The filing does not explain the difference; forfeitures after June, awards outside the plan or a later top-up would each close it, and the September-quarter report should show which.

07 Nasdaq, the auditor, and the register

The listing question, told accurately

There was one listing deficiency in the last eighteen months, and it was not what a reader might assume. The notice of January 12, 2026 was for failure to hold an annual meeting within twelve months of the financial year end. It was not a minimum bid price deficiency, not a shareholders’ equity deficiency, not a market value deficiency and not a late-filing deficiency.

The company submitted a remediation plan, it was accepted, and an extension was granted to March 23, 2026, the date of the scheduled meeting. The meeting was held on March 23, 2026 and its results were filed the next day. The deficiency is therefore cured in fact.

The auditor change

On November 20, 2025 the audit committee appointed Weinberg & Company for the 2025 financial year; the previous auditor, Salberg & Company, was dismissed around November 24. The facts that matter in an auditor change are the ones about disagreement, and here they are clean: no disagreements and no reportable events over the two prior years and the interim period, stated for both the departing auditor and the consultations with the incoming one, and the departing auditor confirmed agreement in its own letter to the regulator.

Three qualifications belong beside that. First, Salberg’s opinions on both 2024 and 2023 carried going-concern explanatory paragraphs — so the going-concern language did not arrive with the new auditor. Second, the incoming auditor was already the auditor of the clinic business NRx acquired, which is a relationship worth knowing even though it is disclosed and unremarkable in itself. Third, there is no Item 4.02 anywhere in the record — no restatement, no non-reliance determination — and only one late filing, a single notification in August 2025 for the June quarter, filed four days later. There is no late annual report in the record. Shareholders ratified the new auditor in March 2026 with 17,351,069 votes in favour against 256,023.

Who owns it, and the short position

The ownership record here is thin, and the thinness is itself the finding. There are only four beneficial-ownership filings in the whole recent record and no Schedule 13D at all.

No aggregate institutional-ownership figure can be sourced from the company’s own filings, because quarterly manager reports are filed by the managers, not the issuer. Any percentage a reader sees quoted for institutional ownership of this stock does not come from a primary filing by the company, and this page publishes none.

At the September 15 settlement the position was about 22.6 per cent of the 43,295,767 shares outstanding, a calculation on the August cover count. The 16.27 days-to-cover figure is what makes it unusual: the position grew while the reported average daily volume in the same series fell by roughly 60 per cent from its August peak, so it would now take more than three trading weeks of normal volume to close. On FINRA’s calendar the September 30 settlement is published on October 9, 2026 and the October 15 settlement on October 26, so there is no September 30 figure yet. Nasdaq short interest →

Insiders, and what has not happened

No Form 4 has been filed since February 17, 2026. That one reported a grant of 100,000 options to the chief operating officer, dated January 27 with an exercise price of $2.39 — an award, not a purchase — and it was filed together with his Form 3, both after their deadlines. Across the whole period covered here, no insider has bought a single share on the open market. Four reports filed in September 2025 covered transactions dated the previous April, five months late against the two-business-day requirement. And the 398,353 restricted units granted to three executive officers on July 28, 2026 do not appear in any Form 4 as of October 5, 2026; no Form 3 has been filed for the new chief commercial officer either.

Governance, in brief

The interim chief executive was made permanent on January 4, 2026, having held the role since October 2024 and having previously led the company until 2022; he is also chairman, and is paid through a consulting agreement at $575,000 a year with a performance bonus whose minimum target is $250,000. The July grant note in the 10-Q lists an “Executive Chairman” and a “Chief Executive Officer” as separate recipients, although the company’s releases through September still describe Dr. Javitt as Chairman and CEO and no Form 8-K reports a change in either role. A chief operating officer was engaged from December 2025 and formalised in January 2026, with 100,000 options at $2.39 carrying acceleration on a change in control. A first chief commercial officer was appointed in 2026. The former chief business officer, Matthew Duffy, left effective December 31, 2025, having taken $371,000 in consulting fees and no salary during the year. The Naples clinic is leased from an entity of Dura’s former owner, who became a director and minority shareholder after the acquisition. Severance arrangements are individual rather than a general plan, capped at six months of salary for both the finance and operations officers. There is no securities class action against the company in the record, and no disclosed regulatory investigation.

One internal inconsistency is worth printing because a reader checking the filing will hit it. The litigation note states that “there are no material pending or threatened legal proceedings at this time” — and the same note, in the same document, describes the Kadima arbitration and lawsuit as pending. Both sentences are in the June quarterly report.

08 Catalyst map, red flags and the Merlintrader bottom line

The catalyst map

The useful thing about this map is the division running down the middle of it. One event has a date set by an agency. Everything else is a company statement, and several of those have already missed dates they set themselves.

Red flags

• The company states that substantial doubt exists about its ability to continue as a going concern, and the auditor added the same paragraph to the 2025 annual accounts.
• Liabilities exceed assets. Total liabilities $34.330 million against total assets $34.160 million, leaving a stockholders’ deficit, with working capital negative $4.7 million.
• The two runway statements published with the same results do not agree — “through at least the first quarter of 2027” in the filing against “at least one year” in the release.
• The share count rose 36 per cent in under eight months, from 31,734,333 to 43,295,767, with the at-the-market programme producing more than twice as much in the June quarter as in the March quarter.
• The at-the-market selling has been done cheaply. 4,812,972 shares over the half at an implied average gross price near $2.51, including a block of 1,195,290 shares around $1.80 between February and March — against the $3.50 of the June underwritten offering. The remaining capacity is not disclosed.
• Short interest is about 22.6 per cent of the share count at 16.27 days to cover, up from 2.46 million shares in February to 9.77 million in September.
• No insider has bought on the open market at any point, no Form 4 has been filed in six months, and the July 2026 executive grants of 398,353 units and 25,000 options are not yet reported on any Form 4.
• The DARPA contract announced on September 30 has not reached a filing. The July filing said the selection was not an award; the September press release announces “an initial $11,177,757.27 contract”, but no Form 8-K, contract number or period of performance is public as of October 5.
• The company’s description of ketamine as a drug on the current shortage list is not matched by the FDA’s public database, which lists ketamine injection only under discontinuations.
• The branded application has been guided three times and missed three times — the fourth quarter of 2025 in the November 2025 10-Q, the second quarter of 2026 in the annual report, then the end of the third quarter of 2026 in the August filing, a quarter that closed on September 30 with no submission announced. The company attributes the delay to the same vial issue that held up the generic application.
• The defence trial is past its registered start date and still not recruiting.
• The clinic spin-off has been announced since 2024 with no registration statement, no date, and a risk factor reserving the right to abandon it. The subsidiary is not an SEC registrant.
• One clinic acquisition collapsed into arbitration and litigation, which the company says it does not expect to have a material adverse effect. The other three clinic transactions — Dura, a minority stake in Cohen & Associates and the Boca Raton clinic assets — did close.
• The litigation note contradicts itself, stating there are no material pending proceedings while describing pending proceedings in the same note.
• The auditor was changed in November 2025 — cleanly, with no disagreements, but the incoming firm was already the auditor of the acquired clinic business.
• 42.9 per cent of votes cast opposed the equity-plan amendment at the March 2026 meeting, which raised the annual evergreen from 1 per cent to as much as 5 per cent of fully diluted shares.
• A warrant tranche was repriced to $1.65 under full-ratchet clauses, adding 1,870,960 shares to the underlying at a fair-value cost of $5.7 million.
• No formal Nasdaq compliance confirmation is on the record for the 2026 deficiency, even though the condition that caused it was satisfied in March.
• A 1-for-10 reverse share combination took effect on April 2, 2024, so any unadjusted price or per-share figure from before it is on a different basis.

What to watch, in order

1. November 10, and whether a second major amendment appears before it. A minor amendment keeps the clock short; a major one restarts it. That single classification matters more than anything the company says about launch plans.
2. The September-quarter report, due by November 16. Three numbers in it: the September 30 cash position, the quarterly burn after the June raise, and whether the runway sentence is restated and in which of the two forms.
3. At-the-market activity since June 30. It produced $8.404 million in the June quarter alone, and the remaining capacity is not disclosed, so the quarterly report is the only place to see it.
4. Whether the branded application is actually filed. Started in late 2024, guided for the fourth quarter of 2025, then the second quarter of 2026, then the end of the third quarter of 2026, and missed all three times. A filed application creates a review clock; an intention does not.
5. Whether the DARPA contract reaches a filing. An initial $11.18 million of non-dilutive funding would change the funding picture for the defence trial. A Form 8-K with the contract, or its terms in the November quarterly report, is the confirmation the September 30 release does not yet provide.
6. Whether SPARC-TMS starts recruiting. The registry is the fastest public check: the September 30 release says enrolment will now begin, the September 25 registry record still reads not yet recruiting, and the estimated start date has already passed.
7. Clinic revenue growth, and whether HOPE is ever broken out. $1.101 million in the June quarter is the only operating revenue the company has, and the segment note shows it profitable at the operating line — $98 thousand of operating income in the quarter — which is the one part of the business that currently pays for itself.
8. Any Form 4 at all. Six months of silence through a capital raise and a regulatory milestone is itself information, and the July grants are already overdue for reporting.
9. Short interest at the next settlements, the September 30 figure on October 9 and the October 15 figure on October 26. At 16.27 days to cover, the position is large relative to the liquidity available to close it in either direction.

Merlintrader bottom line

The useful way to read this company is to sort its announcements by who issued them. One document in the file was written by the FDA and carries a date: the acknowledgement letter setting a GDUFA goal date of November 10, 2026 on the generic application for preservative-free intravenous ketamine. By the company’s account, that application cleared its first review cycle with no major deficiency relating to the drug itself, and the single remaining objection — a twist-off cap on a luer lock vial — was downgraded from major to minor on September 15 after the company produced manufacturer certifications, which puts the review on a three-month clock rather than the six to ten months of a major amendment. The company calls ketamine a shortage drug; the FDA’s public database does not list it as one. That is a legible, near, non-binary catalyst of a kind this sector rarely offers, and it is the reason the page exists.

Everything else in the file was written by the company, and several of those items are not what they are usually called. The DARPA item went from a July selection that the filing said was not an award to a September 30 press release announcing an initial contract of $11.18 million; until a filing or a government record confirms the contract and its terms, it is the company’s announcement rather than a documented award. The branded application for the same molecule was guided for the second quarter of 2026 and has not been filed. The defence trial is past its own registered start date and not recruiting. The clinic spin-off has been announced since 2024 with no registration statement, no date, and a risk factor reserving the right to abandon it. None of that makes the generic application less real; all of it is relevant to how much weight to put on the next announcement.

The balance sheet is where a reader should be most careful, because it is the constraint on everything above. The company itself says substantial doubt exists about going concern. At June 30 liabilities of $34.330 million exceeded assets, leaving a stockholders’ deficit and negative working capital of $4.7 million, on $26.686 million of cash and a burn near $3.0 million a month. The real debt is trivial — a $412 thousand insurance loan, with the old lender debt converted away in December 2025 — and most of the liability total is a non-cash warrant derivative of $18.550 million, so the balance sheet is less alarming than the headline total suggests. But the share count rose 36 per cent in under eight months, and the at-the-market programme is open against a $150 million shelf whose remaining capacity is not disclosed — having already sold 4,812,972 shares over the half at an implied average gross price near $2.51, well under the $3.50 of the June underwritten offering. The two runway sentences published three days apart do not agree with each other.

Which sets up the only question that matters between now and mid-November, and it is not a scientific one. Does the approval arrive before the company needs to raise again, or after? An approval reached with the current share count is a different outcome for existing holders than the same approval reached after further issuance at these levels — and with short interest at 22.6 per cent of the count and 16.27 days to cover, the market has taken a large position on that sequencing. The two documents that will answer it are the FDA’s action and the September-quarter report, due on November 10 and by November 16 respectively.

None of the above is a recommendation, a target, or a view on the share price. It is a description of what is documented, what is guided, what is calculated and labelled as such, and what the record does not say.

Method, and what is not verified. Balance-sheet, cash-flow and expense figures, the share counts, the warrant and option tables, the going-concern conclusion and the runway sentence come from the Form 10-Q for the quarter ended June 30, 2026, filed August 14, 2026; prior-year figures from the annual report filed March 23, 2026. Regulatory facts come from the Forms 8-K read with their exhibits rather than their cover pages, which matters here because the limiting language on the DARPA selection is in the body of the filing and the goal date is in the exhibit. Market figures come from the exchange’s own daily series at the October 2, 2026 close. Where a company statement and an agency record disagree, as on the shortage list, both are given and attributed. The market value, the fully diluted count, the June-quarter burn, the share-count increase, the short interest as a percentage of shares outstanding and the implied current size of the unamended B Group stake are calculations from published data and are labelled as calculations where they appear.

The following are not established on the public record and are not asserted here: any September 30, 2026 cash figure or burn rate; any approval, tentative approval or refusal on the abbreviated application, November 10 being a goal date rather than a decision; any formal acceptance of the manufacturer certifications requested on August 6, the September 15 reclassification being strong evidence but not a statement of acceptance; any filed DARPA contract or its terms, the July 21 letter being a selection for negotiation that the filing says is not an award and the September 30 award being, so far, a press-release announcement; any filed branded application or PDUFA date for NRX-100 or NRX-101; any registration statement, date or structure for the clinic spin-off, and the clinic subsidiary is not an SEC registrant; any valuation figure for the clinic business, although its revenue and operating result are disclosed in the segment note; any completion of the January 2025 preferred-funding agreement for the clinic subsidiary, with its royalty on drug sales; the remaining at-the-market capacity, which the company does not disclose — the share counts sold are disclosed, so the average price quoted here is a calculation from them; the non-affiliate float at any 2026 date; any written Nasdaq confirmation that compliance was regained for the 2026 deficiency; any aggregate institutional-ownership percentage, which cannot be sourced from the issuer’s own filings; any current position for The B Group after August 2025; and any analyst rating, consensus or price target, none of which is verifiable from a primary source and none of which this page publishes. There is also no securities class action against the company and no disclosed regulatory investigation in the record. Third-party screen data for float, short float and retail sentiment has not been re-verified and is not reproduced.