Nyxoah ($NYXH) Stock Hub 2026: ACCCESS Meets Both Primary Endpoints With a 77.2% Responder Rate, Opening the U.S. Door to CCC Patients

01 Executive summary: a second pivotal win for a company that is already selling

Nyxoah SA is a Belgian medical technology company with one product, the Genio system, a bilateral hypoglossal nerve stimulator for moderate to severe obstructive sleep apnea. The device has carried a CE mark since 2019, a European indication for complete concentric collapse of the soft palate since 2021, and a U.S. premarket approval since August 8, 2025 for patients with an apnea-hypopnea index between 15 and 65. The company listed on Euronext Brussels in September 2020 and on Nasdaq in July 2021, and files with the SEC as a foreign private issuer on Forms 6-K and 20-F.

On September 2, 2026 the company reported the topline of ACCCESS, its second U.S. pivotal study. ACCCESS enrolled patients with complete concentric collapse, the anatomical pattern that today excludes a patient from unilateral hypoglossal nerve stimulation in the United States and that Genio’s own U.S. labelling lists in its warning section. At twelve months, 77.2% of patients were AHI responders under the Sher criteria and 87.7% were ODI responders, both with p<0.001 on a Full Analysis Set of 47 patients, and no device-related serious adverse event was recorded. The company is preparing a PMA supplement to extend the U.S. indication to this population, under the Breakthrough Device Designation the FDA granted to the programme, and will present the complete dataset at the International Surgical Sleep Society meeting on October 15-16, 2026.

The clinical news lands on a company that is no longer pre-commercial. Net revenue was EUR 7.7 million in the second quarter of 2026, of which EUR 5.2 million from the United States, up 22% on the first quarter. The company had 180 active high-volume U.S. accounts, 262 trained surgeons and 427 patients in prior authorisation at the end of June, and it kept its 2026 guidance of EUR 36-40 million of revenue. It also raised $95 million of equity and drew $15 million from the European Investment Bank in June, taking cash and financial assets to EUR 97.8 million on June 30, 2026 from EUR 25.9 million on March 31.

What the file still carries: an operating loss of EUR 41.1 million in the first half, EUR 58.1 million of financial debt including a EUR 16.8 million convertible bond that converts at EUR 1.48 and can be serviced in shares, a share count that went from 44.7 million before the June offering to 101.8 million on August 25, a consolidated patent case with Inspire Medical Systems in Delaware, and a chief executive transition announced on June 4, 2026 with the current CEO staying on until a U.S.-based successor is appointed.

02 ACCCESS: what the September 2, 2026 release actually says

ACCCESS is a multicentre, prospective, open-label pivotal study run in the United States under an investigational device exemption the FDA approved in July 2022. It enrolled adults with moderate to severe obstructive sleep apnea and complete concentric collapse of the soft palate who had failed, did not tolerate or refused positive airway pressure. The co-primary efficacy endpoints are the AHI responder rate under the Sher criteria and the ODI responder rate, both measured on a twelve-month polysomnography, with patients followed for three years. On ClinicalTrials.gov the study is NCT05592002, started on December 28, 2022, across sixteen sites, with a registry enrollment figure of 124 and a primary completion estimated for September 2026.

Three things in the release deserve a plain reading. First, the two responder rates are on the Full Analysis Set, not on a per-protocol set, and the company says the analysis follows the protocol as approved by the FDA; that is the population a reviewer will look at first. Second, 47 patients is a small number. The August 2025 decision to close enrollment ahead of the 106-patient maximum was presented at the time as confidence in the statistical power, and the p-values bear that out, but confidence intervals, the number of patients who were implanted but did not reach the twelve-month sleep test, and the imputation sensitivity analyses are not in the release. Third, “zero device-related serious adverse events” is a statement about the serious category attributed to the device; procedure-related events and non-serious events will be in the ISSS dataset.

03 Genio and the complete concentric collapse problem

Hypoglossal nerve stimulation treats obstructive sleep apnea by stimulating the nerve that controls the tongue, so that the tongue moves forward during sleep and keeps the airway open. The therapy exists in the United States since 2014 through Inspire Medical Systems, whose system stimulates one side of the nerve with an implanted pulse generator, a sensing lead and a stimulation lead. Patient selection for that therapy depends on a drug-induced sleep endoscopy: if the soft palate collapses in a complete concentric pattern, the patient is contraindicated, because unilateral tongue protrusion does not resolve that pattern of obstruction.

Genio is built differently. It is a single implant placed under the chin that stimulates both branches of the hypoglossal nerve, with no implanted battery and no leads; power and control come from an external activation chip worn on a disposable patch under the chin during sleep. The company’s argument, and the reason the CCC population was worth a dedicated pivotal study, is that bilateral stimulation moves the whole tongue base and can open an airway that collapses concentrically. The BETTER SLEEP study supported the European CCC indication in 2021; ACCCESS is the U.S. equivalent.

The commercial weight of the CCC population is not something the company quantified in the release. The co-principal investigator, Colin Huntley of Thomas Jefferson University, said in the release that “a significant number of patients with OSA are found to have CCC”. Estimates of the share of HGNS candidates screened out for CCC circulate in the sleep-surgery literature, but no company figure is on the record, so no figure is given.

04 Regulatory path: what has been said, and what has not

The company’s stated next step is a premarket approval supplement to the existing PMA, under the Breakthrough Device Designation the FDA granted to Genio for the CCC population. A PMA supplement is the route for a change to the indications for use of an approved class III device. The FDA reviews it against the clinical data submitted; for an indication expansion of this kind, the review can include an advisory panel at the agency’s discretion.

05 DREAM and the existing U.S. label

The U.S. approval of August 8, 2025 rests on DREAM, the first U.S. pivotal study, NCT03868618, which started in October 2020 across 21 sites with a registry enrollment of 115. DREAM met its primary and secondary endpoints with an AHI responder rate of 63.5%, an ODI responder rate of 71.3%, a median AHI reduction of 70.8% and 82.0% of subjects reaching an AHI below 15. The study also measured position-specific outcomes: a 66.6% median AHI reduction in the supine position against 71.0% non-supine, which the company presents as evidence that the therapy works regardless of sleeping position.

The approved indication is a subset of adults with moderate to severe obstructive sleep apnea, with an AHI of at least 15 and at most 65. The company’s own description of the label says that the treatment of CCC patients is included in the warning section of the indications for use, which is exactly the boundary ACCCESS is designed to move. Genio is a prescription-only device in the United States.

Two smaller points from the same file. To date more than 1,000 patients have been treated with Genio in Europe, according to the June 4, 2026 6-K, and in Germany the company reports a 25% market share within 24 months of launch and 53% implant growth in 2025 over 2024. The same filing states that the collaboration agreement with Cochlear Limited, which dates from November 2018 and covered development work on the implant, can be considered ended, with non-exclusive cross-licences remaining in place.

06 The U.S. launch in numbers, one year after approval

The launch strategy is a direct sales organisation focused on the top 400 hypoglossal nerve stimulation accounts, which the company estimates account for 70% to 75% of all HGNS procedures in the United States, with a narrower priority list of 200 high-volume accounts for the 40 territory managers hired through the first half of 2026. The company reports the leading indicators every quarter.

Sources: Q4 and FY 2025 release of March 19, 2026; 6-K of June 4, 2026 for the first-quarter indicators; preliminary Q2 release of July 7, 2026 and Q2 release of August 5, 2026. The Q1 2026 U.S. figure is a Merlintrader calculation from the reported EUR 5.2 million and 22% sequential growth; the Q1 2026 global figure is the H1 total minus Q2. “n.a.” means the company did not report the metric for that date.

The reading is straightforward: every leading indicator roughly doubled between December and June, and revenue followed with a lag because a patient has to be authorised, scheduled and implanted before a system is invoiced. The 427 patients in prior authorisation entering the third quarter are the best forward indicator the company gives, and the July 8 investor day in New York, hosted at Bank of America, added qualitative support from ENT and sleep physicians rather than new numbers.

07 Reimbursement: the C-code, the CPT question and the 2027 proposals

Genio has a dedicated Medicare C-code, C8011, and is billed by commercial payers under existing hypoglossal nerve stimulation CPT codes. On the July 7, 2026 preliminary release, CMS had proposed to raise the hospital outpatient payment for the Genio procedure from $31,526 to $35,414 for 2027, an increase of $3,888 or 12%, and the ambulatory surgical centre payment from $27,563 to $31,722, an increase of $4,159 or 15%. The company describes these as among the strongest proposed increases within APC 5465, which the release labels “Level 5 Neuromodulation”. Proposed rules become final in the autumn, typically in November, with effect from January 1.

The open item is coding. The company refers to upcoming CPT editorial panel meetings, involving the American Academy of Otolaryngology-Head and Neck Surgery and industry, to discuss comprehensive HGNS coding. A change in how HGNS procedures are coded would affect both Genio and Inspire; the company’s reimbursement expert group at the July investor day argued that Genio is positioned across the coding scenarios. Nothing is decided, and the outcome of a CPT panel is not a date the company controls.

08 Clinical map: every Genio study on the registry

Source: ClinicalTrials.gov records read on September 3, 2026, and company releases. Registry enrollment figures include screened and roll-in subjects and do not match the analysis sets in the company’s releases; the ACCCESS release states the analysis set of 47 and the planned maximum of 106.

09 Financials: the second quarter and the first half of 2026

The company reports under IFRS in euros. The figures are from the unaudited consolidated statement of loss in the August 5, 2026 release, filed with the SEC as exhibit 99.1 to a 6-K.

The operating loss is almost flat year on year, EUR 41.1 million against EUR 40.5 million for the half, because the revenue increase of EUR 11.6 million and the EUR 0.7 million cut in research spending were absorbed by EUR 7.9 million of additional selling, general and administrative expense, which is the cost of the U.S. sales organisation. Below the operating line, the second quarter shows EUR 13.1 million of financial expense against EUR 3.3 million a year earlier; the half-year financial expense is EUR 9.4 million, which means the first quarter carried a financial gain. The company’s release does not break the item down, and the June convertible-bond terms, the EUR-USD moves on the dollar offering proceeds and the bond’s fair-value accounting are the candidates. The 20-F and the interim report are the documents that will say which.

Guidance for 2026 is unchanged on revenue and margin and moved by EUR 1 million on total operating expenses: net revenue of EUR 36-40 million, gross margin of 60-62%, total operating expenses of EUR 99-102 million, up from the prior range by about EUR 1 million for a one-time share-based compensation charge of about EUR 0.9 million tied to the repricing of employee equity incentives in the second quarter, and non-GAAP cash operating expenses of EUR 88-90 million. First-half revenue of EUR 14.0 million means the second half needs EUR 22-26 million, or EUR 11-13 million a quarter, against EUR 7.7 million in Q2.

10 Cash, burn and runway: what the June financing bought

On March 31, 2026 the company had EUR 25.9 million of cash, cash equivalents and financial assets and an accumulated deficit of EUR 321.7 million, and its own 6-K of June 4 said that, on the operating plan and the year-end cash plus the second EIB tranche, the runway extended into the third quarter of 2026. That is the position from which the company launched the offering the same evening.

The cash and financial assets line rose by EUR 49.8 million in six months while the company lost EUR 41.1 million at the operating level, which is consistent with roughly EUR 81.7 million of gross equity proceeds at the EUR 1.48 offering price, EUR 13.8 million from the EIB, and roughly EUR 45 million absorbed in the half by operations, underwriting discounts and offering costs, working capital and debt service. Financial debt rose by EUR 17.5 million, the EIB tranche plus the movement in the convertible. The company has not restated a runway in months or quarters since the financing; the chief executive said on July 7 that “the financial overhang is now behind us”. On the reported numbers, EUR 97.8 million against a cash operating expense guidance of EUR 88-90 million for the full year, roughly EUR 45 million of which falls in the second half, and against EUR 23.1 million of debt classified as current, the arithmetic supports operations through 2027 only if revenue keeps growing at the guided pace; a dated runway statement from the company would replace that calculation.

11 Capital structure and dilution: 101.8 million shares, a convertible and an ATM

Nyxoah has raised EUR 332.5 million of equity since inception, on the company’s own count in the June 4, 2026 6-K: EUR 103.6 million in the Euronext IPO of September 2020, EUR 75.0 million net in the Nasdaq IPO of July 2021, $32.8 million under the first Cantor Fitzgerald ATM through December 2024, EUR 18.9 million in 2023, EUR 69.7 million in 2024 and EUR 21.9 million in 2025. The June 2026 offering is the largest single raise since the IPOs and the most dilutive.

The share count tells the story in three notices. On June 10, 2026, after the offering, the company reported 99,926,284 shares, which means 44,693,726 shares existed before it. On June 30 the count was 100,072,815 after 146,531 restricted share units vested. On August 25 it was 101,840,217 after the convertible holder’s 1,767,402 shares. The share capital is EUR 7,092,826.33. Basic loss per share for the half was EUR 0.976 on the weighted count; on the current count the same loss would be about EUR 0.48, which is the mechanical effect of doubling the denominator mid-year.

12 Ownership: AIGH above 10%, directors in the offering, ResMed below 2%

Belgian transparency law obliges holders to notify when they cross thresholds, and the company publishes those notifications. Two arrived in June 2026. Entities controlled by Orin Hirchman, through AIGH Capital Management LLC, notified 14,534,880 voting rights on June 10, 2026, or 14.55% of the 99,926,284 shares then outstanding, split between AIGH Investment Partners, LP with 10.56% and the WVP Emerging Manager Onshore Fund AIGH Series with 3.98%; the threshold crossed was 10%, on the day the offering shares were issued. ResMed Inc. notified a passive downward crossing of the 3% threshold on the same date, holding 1,943,828 shares or 1.95%, up from 1,727,864 shares at its previous notification: the holding grew, but the enlarged share count took it below the 3% threshold. The notification does not say when or where the additional shares were acquired.

The pricing release names three directors who had expressed interest in taking shares in the offering: the chairman Robert Taub, through Robelga SRL, and the independent directors Jürgen Hambrecht and Kevin Rakin. The board applied the Belgian related-party procedure, a committee of three independent directors, Rita Johnson-Mills, Virginia Kirby and Wildman Ventures represented by Daniel Wildman, advised that the participation was on market terms, and the board did not deviate from the advice. The allocations to the three were not disclosed in the releases read.

As a foreign private issuer Nyxoah’s insiders do not file Forms 4, so there is no U.S. record of open-market director purchases to reproduce. The insider percentage on the Finviz export, 20.18%, is an aggregator figure that presumably includes the chairman’s historical stake; the transparency register on the company’s investor site is the primary source for any individual holding above 3%.

13 Management and governance: a CEO search in the middle of the launch

On June 4, 2026, the same day it launched the offering, Nyxoah announced that its chief executive, Olivier Taelman, and the board had decided to transition leadership to a U.S.-based chief executive, and that the board had formally launched a search. Taelman, who has led the company for seven years, remains in charge of daily operations during the transition and will support the onboarding of the successor. No successor has been named as of September 3, 2026, and every release since, including the ACCCESS release, quotes Taelman as chief executive officer.

The chief financial officer is John Landry, who signs the 6-Ks and is the investor relations contact; Rémi Renard is head of investor relations and corporate communication. The chairman is Robert Taub, the company’s founder, through Robelga SRL. The June 10, 2026 annual meeting agenda proposed to reappoint Robelga SRL, Jürgen Hambrecht, Kevin Rakin, Rita Johnson-Mills, Virginia Kirby and Pierre Gianello until the 2027 meeting and to appoint Daniel Wildman as independent director; Wildman Ventures, represented by him, already sat on the June 4 committee of independent directors that reviewed the related-party participation in the offering, and the filings read do not say when he first joined the board. The extraordinary meeting proposed to renew the authorised capital for five years. The company has not published a 6-K with the voting results in the filings read; the outcome of the meeting should be checked on the investor site’s shareholder-information page.

14 Litigation with Inspire Medical Systems

The company’s June 4, 2026 6-K sets out the state of the patent dispute with Inspire, which is the only material litigation disclosed. Inspire sued Nyxoah SA and Nyxoah, Inc. in the District of Delaware on May 30, 2025, alleging that Genio infringes U.S. patents 10,898,709, 11,806,526 and 11,850,424; Nyxoah counterclaimed for a declaration of non-infringement and invalidity. Nyxoah sued Inspire in the same court on September 15, 2025, alleging that the Inspire IV and Inspire V systems infringe U.S. patents 8,700,183, 9,415,215 and 9,415,216; Inspire moved to dismiss part of that complaint, Nyxoah amended, and Inspire renewed the motion. On May 15, 2026 the court consolidated the two cases and ordered a joint schedule by June 18, 2026.

On December 1, 2025 Nyxoah filed two actions in the Unified Patent Court in Munich against Inspire and Inspire Medical Systems Europe GmbH on European patents EP 2 760 528 B1 and EP 2 760 534 B1, seeking damages and an injunction. On December 18, 2025 Nyxoah petitioned the U.S. Patent and Trademark Office for inter partes review of the three Inspire patents; on April 14, 2026 the petitions were denied institution, which means the validity of the Inspire patents will be argued in the district court rather than before the Patent Trial and Appeal Board.

15 Competition and the market Genio sells into

In the United States the reference competitor is Inspire Medical Systems, which has sold unilateral hypoglossal nerve stimulation since 2014 and defines the accounts, the surgeons and the reimbursement codes that Genio is now entering. The company’s own framing is that Genio competes on three product attributes, bilateral stimulation, no implanted battery and no leads, and on one clinical attribute, the CCC population, which ACCCESS is meant to turn into an on-label advantage. Inspire’s results, its share of the 400 target accounts and its response to a Genio CCC indication are not in Nyxoah’s filings and are not reproduced from secondary sources.

The demand side is set out in the June 2026 6-K: the company expects U.S. obstructive sleep apnea prevalence to grow by about 35% by 2050, cites an approximately 11% higher rate of positive airway pressure initiation among GLP-1 patients, and expects the global hypoglossal nerve stimulation market to grow approximately 15.8% from 2026 to 2033, a figure the filing gives without saying whether it is cumulative or annual. The GLP-1 point is the company’s answer to the question every device investor asks about weight-loss drugs and sleep apnea: on its reading, the drugs bring more patients into diagnosis and treatment rather than removing them from the surgical pool. That is a company view, and the 2033 growth rate is a company forecast.

16 Market data, valuation arithmetic and short interest

The Nasdaq close on September 2, 2026, the day of the ACCCESS release and before it, was $1.51 on Marketstack end-of-day data. Applied to the 101,840,217 shares of August 25, 2026, that gives a basic market capitalisation of about $153.8 million, a Merlintrader calculation. The June 2026 offering was priced at $1.72; the convertible converts at EUR 1.48.

Enterprise value is not computed here because the market capitalisation is in dollars and the balance sheet in euros, and converting one into the other on a chosen day would present a rate as if it were a fact. Two observations hold in either currency: the company trades at a basic equity value that is a low single-digit multiple of the 2026 revenue guidance, and its net cash of about EUR 40 million covers less than a year of the guided cash operating expenses net of gross profit. Short interest below 1% of the float on the Nasdaq line says the stock is not a crowded short; the Euronext Brussels line, where a large part of the float sits, is not covered by U.S. short-interest data.

17 Analysts

No analyst note with a verified house and publication date was available at the September 3, 2026 cut-off, and the company’s investor site did not return its analyst-coverage page to a plain request. Aggregator screens show a consensus price target and a rating average for the stock, but a table built on numbers whose authors and dates are unknown would be a table of rumours, so it is not reproduced. When a note is verifiable it will be listed with the house, the date and the target; until then, the offering price of $1.72 set on June 5, 2026 by BofA Securities as lead bookrunner is the only third-party price on the record.

18 Retail sentiment: extremely bullish reading on the morning after the release

The Stocktwits reading was taken on September 3, 2026 at about 2:20 am ET, roughly ten hours after the ACCCESS release. The normalised sentiment score was 87 out of 100, which the platform labels extremely bullish, and the message-volume score was 81 out of 100, labelled extremely high, on a stream with 431 watchers. The platform did not expose a bullish-to-bearish split of tagged messages at the snapshot, so no split bar is shown. The stream itself was dominated by extended-session gainer lists and watchlist posts that grouped the ticker with unrelated names, plus a handful of comments on price levels; almost none of the posts discussed the study.

19 Merlintrader Health Score

The score weighs five pillars with fixed weights, balance sheet and runway 30%, catalysts 30%, dilution 20%, liquidity 10% and execution 10%, and describes how robust or fragile the company looks over the next twelve to eighteen months. It is an editorial judgement on the documents read at the September 3, 2026 cut-off.

20 Catalyst table

Every date marked as estimated is a Merlintrader inference from the company’s prior pattern and is not company guidance. Dates the company has confirmed are marked as such.

21 Scenario framework, not a forecast

22 Red flags and thesis-breakers

• A PMA supplement that does not get filed on the timeline the market assumes. The company said “preparing”; every month without a filing date pushes the label expansion further out.
• An FDA request for additional data or an advisory panel on a 47-patient analysis set, which would lengthen the review and revive the question of why enrollment stopped at less than half the planned maximum.
• Financial expense that turns out to be cash. EUR 13.1 million in one quarter needs an explanation in the interim report; if the convertible’s mechanics are the reason, the cost of that capital is far above its coupon.
• Convertible settlement in shares at falling prices. The relevant share-settlement price was EUR 1.25 on August 18 against a EUR 1.48 conversion price; each notice showing a lower settlement price and more shares issued is a red flag on its own.
• Use of the $50 million ATM before a value-inflecting event, or a further offering below $1.72, the price of the June raise that was described as removing the financing overhang.
• Third-quarter revenue below EUR 10 million, which would put the EUR 36-40 million full-year guidance out of reach and question the conversion of the 427 patients in prior authorisation.
• A CEO transition that drags past the end of 2026, or a successor without U.S. device commercial experience.
• An adverse ruling in Delaware, in particular any preliminary or permanent injunction against Genio in the United States, now that the inter partes review route is closed.
• A CPT coding change that reduces the economics of HGNS procedures for hospitals or surgeons, or a final CMS rule that abandons the proposed 2027 increases.
• Going-concern language in the 2026 annual report, which would signal that the June financing and revenue growth together are not judged sufficient for twelve months.

23 Evergreen monitoring checklist

24 Bottom line

Nyxoah is a rare shape in small-cap medtech: a company that has already cleared the FDA once, is selling into the accounts that matter with numbers that doubled in six months, and has just produced a second pivotal result in a population its main competitor cannot treat. The ACCCESS topline of September 2, 2026, 77.2% AHI responders, 87.7% ODI responders, no device-related serious adverse events in 47 patients, is a clean clinical statement, and the company has a defined regulatory route to turn it into a U.S. indication.

The financial shape is the other half of the file. The company lost EUR 41.1 million at the operating level in six months on EUR 14.0 million of revenue, carries EUR 58.1 million of debt including a convertible that converts at EUR 1.48 and is partly serviced in shares, has doubled its share count in 2026, and is run by a chief executive who has announced his own succession. The June financing bought time and a U.S. sales force; it did not buy profitability, and the 2026 guidance of EUR 36-40 million of revenue against EUR 88-90 million of cash operating expenses says how far there is to go.

What decides the next twelve months is a sequence, not a single event: the full ACCCESS dataset on October 15-16, the PMA supplement filing, third-quarter revenue against the pace the guidance implies, the CMS final rule, the name of the next chief executive, and the shape of each share-count notice as the convertible is serviced. The clinical leg is now the strongest leg of the story. The others are the ones to watch.

Related research on Merlintrader

• Top Ten Biotech Stocks Right Now
• Free Biotech Catalyst Calendar: FDA and PDUFA dates