This is a placement study, not a performance study. The variable being tested is where the test sits in the pathway, GP room versus memory clinic, because placement is what determines volume. UK dementia diagnosis is throughput-constrained, and PET/CSF capacity is the choke point. If the assay holds up at the GP gate, C2N converts from a specialist-ordered confirmatory test into a primary-care triage tool: a different order of magnitude in units, and an entirely different reimbursement conversation.

• MHRA registration allows sale; it does not create funding. The Accurate Dx evidence is being generated to support health-technology assessment and NHS commissioning. The DAC framing, “healthcare system preparedness”, telegraphs the argument: systems must be ready before anti-amyloid demand arrives, not after.
• Competitive frame: Roche’s Elecsys pTau217 and Fujirebio’s Lumipulse are chasing the same gate. First-mover advantage in a national pathway is unusually sticky, because primary-care pathway change is expensive to undo.
• The number to demand: positive and negative predictive value in an unselected primary-care population. Performance in symptomatic specialist cohorts does not transfer cleanly to lower-prevalence primary care, where PPV falls and false positives flow straight into the specialist queue the test was meant to relieve.

The scientific pitch is elegant: one oral molecule blocking three migraine-provoking neuropeptides at a single receptor, rather than neutralizing one ligand at a time. The de-risking reference is Lundbeck’s anti-PACAP antibody bocunebart, which delivered monthly-migraine-day reductions in the CGRP range, evidence that the PACAP arm of this pathway is drug-worthy. What remains unproven is that receptor-level blockade downstream of three ligands converts to the same effect size, and that an oral small molecule achieves it.

• Portfolio context matters more than the protocol. EVO756 missed in Phase 2b chronic spontaneous urticaria (topline, June 2026). Atopic dermatitis Phase 2b topline is due 2H 2026. Migraine has been promoted from third indication to load-bearing one, and the CSU miss leaves MRGPRX2 target validation as an open question these data must answer.
• The real comparator is not the CGRP antibodies. Evommune cites more than 10M US patients eligible for preventive therapy with roughly 60% untreated with advanced agents, and 65% preferring oral to injectable. But the oral preventive slot is already occupied by the gepants. Atogepant and rimegepant are the benchmark EVO756 must beat on efficacy and tolerability, not eptinezumab.
• Timing: 12 weeks of treatment across ~330 patients puts topline in late 2027 on standard enrollment assumptions, after the AD readout that will already have re-priced the target.

The mechanistic claim is that selectively binding toxic amyloid-beta oligomers while avoiding plaque decouples efficacy from ARIA risk. The interim is consistent with that claim, and consistency is the ceiling of what a blinded safety readout can deliver. With a placebo fraction in the denominator and no unblinding, “no ARIA-E in 136 patients” cannot be attributed to drug with confidence; the company’s own hedge on biomarkers (“potentially reflective of the randomization pattern”) concedes the same point.

• Where the value actually sits is access, not efficacy. If PMN310 pairs class-typical biomarker movement with a clean ARIA profile, the commercial argument is not superiority over Leqembi or Kisunla. It is removal of the serial-MRI monitoring burden, the APOE4 genotyping gate and the homozygote warning that shape prescribing today. Fewer MRIs and a lighter monitoring footprint widen the prescriber base well beyond infusion-capable centres.
• The risk is symmetric. An antibody engineered to avoid plaque may also fail to clear it on PET, and the accelerated-approval precedent in this class is built on amyloid PET reduction as the surrogate. ProMIS will need a biomarker package regulators accept without a plaque-clearance endpoint.
• Binary date: Q1 2027, 12-month unblinded topline with efficacy. Everything before it is safety colour.

The strategically interesting decision here is sequencing. Rather than leading with the implant, Neuvotion is launching the non-invasive stimulation layer first, a cleared wearable that can be sold into rehab units today, and building the BCI on top of it. That converts a decade-long implant timeline into near-term revenue and, more valuably, into a clinical install base and an outcomes dataset the implant programme can later recruit from and reimburse against.

• This is the inverse of the Neuralink/Paradromics posture (implant first, indication later) and much closer to how MicroTransponder built paired-VNS stroke rehab with Vivistim: embed in the rehab workflow, then expand the claim.
• Reimbursement is the gate, not clearance. Post-stroke and SCI rehab devices live or die on whether they map to an existing CPT/HCPCS pathway and DME coverage. Watch for coverage or coding announcements alongside the fall roll-out. Their absence would be the tell.
• Provenance is doing real work in the pitch. A Nature Medicine paper authored at Feinstein gives an early-stage device company clinical credibility it could not otherwise buy. Note this is the second Feinstein-origin item this week, and the third anchored on a Nature Medicine publication.

Every vagus nerve stimulation product on the market today stimulates a cable whose internal human wiring was, until now, largely uncharacterized. That is precisely why VNS dosing has remained empirical and why off-target effects (cough, hoarseness, dyspnoea) are the category’s standing tolerability tax. A fibre-resolution, population-scale atlas is the precondition for fascicle-selective electrodes and for computational dose models that can predict which organ a given contact recruits.

• Who this lifts: SetPoint Medical (whose VNS device for rheumatoid arthritis was FDA-approved in July 2025 and first implanted at Northwell), LivaNova in epilepsy and depression, MicroTransponder in stroke rehab, and the entire non-invasive taVNS field. Open release under SPARC means nobody gets exclusivity. It raises the category floor and pushes differentiation onto electrode design and closed-loop control.
• The strategic read on Feinstein is a standards play. Publishing openly rather than licensing means whoever defines the reference anatomy sets the vocabulary that regulatory submissions in this field will use.
• The signal to watch: the first time FDA references REVA anatomy in VNS pre-submission feedback or guidance. That is the moment an open dataset becomes a de facto regulatory expectation, and legacy non-selective designs start carrying a justification burden.

NfL alone already flagged the pre-symptomatic window in this same cohort back in 2017. The advance here is timing precision, and timing precision is what makes prevention trials financeable. A prevention study in genetically at-risk carriers currently has to enroll broadly and wait; a panel that brackets onset to ±18 months lets a sponsor enrich for imminent converters, shrink sample size and shorten follow-up. That is the same manoeuvre that made pre-symptomatic Alzheimer’s trials viable.

• Direct beneficiaries: Biogen’s tofersen, whose ATLAS study already runs in pre-symptomatic SOD1 carriers with NfL-triggered treatment initiation, so this panel is a refinement of exactly that trigger logic. Also relevant to FUS-targeted antisense and any C9orf72 programme contemplating a prevention design.
• Hold the caveats firmly. 33 converters is a small denominator, the cohort is familial, and generalization to sporadic ALS, roughly 90% of cases, is unproven. There is no analytical validation, no CLIA assay and no regulatory qualification. This is a research signature, not a test.
• The uncomfortable corollary: a panel that predicts onset in a disease with no preventive therapy creates a disclosure and counselling problem before it creates a therapeutic one. Sponsors building prevention programmes will need a genetic-counselling infrastructure argument alongside the clinical one.

Two documents, two different audiences. The guidance is for development teams. The Part 15 hearing is the one market-access teams should calendar, because its agenda (training, monitoring, reimbursement, access, centre capacity) is the delivery-model question that determines whether an approved psychedelic is a product or a programme. Nothing in the guidance resolves the fact that a supervised multi-hour dosing session has no clean reimbursement home.

• The NAM and computational-modelling language is the genuinely new lever. It opens a route to argue abuse potential and hallucinogenic liability from receptor pharmacology rather than clinical experience alone. That structurally advantages purpose-designed non-hallucinogenic compounds, Enveric’s EB-003 among them, over classic psychedelics whose liability is intrinsic to the molecule.
• Backdrop: this lands after April’s Executive Order 14401 directing prioritized review for Breakthrough-designated psychedelics and the National Priority Vouchers issued in the same window, with Compass’s COMP360 the nearest-term filing. Finalizing guidance and scheduling a delivery-model hearing is an agency building the operational layer beneath an approval it now expects to make.
• For Enveric specifically, this is positioning, not acceleration. EB-003 is still in IND-enabling genotoxicity work. The guidance is directionally helpful and changes no timeline.

This is unambiguously a commercial-execution round, not a technology round. Five clearances are already banked and the stated use of proceeds is sales, deployment and market access. That is the correct read of where risk now sits for this company: reimbursement and workflow adoption, not signal quality.

• The economic argument is the strongest in ambulatory neuro-diagnostics right now. Extended-duration home recording converts a capacity-limited inpatient service into an outpatient one, and the payer case (shorter time to diagnosis, fewer inconclusive studies, avoided admissions) is intelligible without generating new evidence.
• Competitive set: Ceribell in the acute and emergency setting, Zeto and incumbent wireless systems, plus traditional ambulatory EEG service bureaux. Epitel’s axis of differentiation is duration, weeks rather than days, which is precisely where the diagnostic-yield argument lives.
• The size is modest for a national scale-up. $26M funds a sales build, not saturation. Expect a strategic partnership or a materially larger round inside 18 months; the paediatric clearance down to age one is the asset most likely to attract a strategic.

The most informative lines in this release are not the clearance. They are the CMS alignment and the banker, in that order. Deliberately mapping the workflow onto existing EEG reimbursement means VoxNeuro is not asking payers to create a new code. It is asking them to recognize a familiar one. For a cognitive SaMD that is the fastest available path, and it is the difference between a cleared product and a billable one.

• Announcing a banker engagement simultaneously with clearance reads as a company monetizing the regulatory de-risking event rather than building a US commercial organization. For acquirers and competitors, cfNI just became a discrete, valuable, purchasable asset with a defined price-forming moment.
• The 18-70 age ceiling is the strategic weakness. The commercially dense cognitive-assessment population, the one attached to the amyloid-therapy pathway, sits above 70. Until the 71-85 expansion clears, cfNI’s addressable use is concussion, TBI, psychiatry and TMS monitoring, not dementia work-up.
• Category context: this is the second EEG/ERP cognitive SaMD clearance in three weeks, after Universal Brain’s UB ERP System on 13 July for psychiatry. Objective brain-function measurement is consolidating into a billable adjunct category, and consolidating categories attract consolidators.

Read this as pharmacodynamics, not efficacy. Three patients, open-label, no control arm, an imaging endpoint, and a within-patient pre/post comparison of microglial PET signal. Translocator-protein PET carries meaningful test-retest variability and is confounded by TSPO genotype. A 26% SUVR change in n=3 is a signal worth pursuing. It is not evidence of clinical benefit in a disease that typically kills within six to ten years.

• Read the cadence, not the number. Tiziana has now issued MSA PET releases at n=2 and again at n=3. Patient-by-patient disclosure from an uncontrolled study is a financing posture, and it establishes an expectation the company will have to keep feeding.
• The asset thesis is real regardless. Intranasal anti-CD3 to induce regulatory T cells is a genuinely differentiated route into neuroinflammation, and the same mechanistic claim now spans na-SPMS, moderate Alzheimer’s and MSA. That breadth is simultaneously the strength (platform) and the weakness (no single indication yet has a controlled dataset).
• The readout that will actually move the asset is the randomized, placebo-controlled na-SPMS Phase 2a. Everything else is supportive colour until that reads out.