This Week in Neurology — Sep 30, 2026

This AudioScholar briefing is generated by artificial intelligence for healthcare professionals and trainees. It is not medical advice.

Welcome to This Week in Neurology. This week we're covering 10 notable papers spanning imaging biomarkers in demyelinating disease, precision diagnosis through genetics and seeding assays, and prevention, prognosis and trial design in cognitive and cerebrovascular neurology. Let's dive in.

We start with imaging, where three papers converge on the question of how much a scan can really tell us. Nature Reviews Neurology carries MAGNIMS consensus recommendations from Cortese and colleagues on the practical use of magnetic resonance imaging in aquaporin-4 positive neuromyelitis optica spectrum disorder and in myelin oligodendrocyte glycoprotein antibody-associated disease, or MOGAD [1]. These are expert recommendations rather than trial evidence, and they set out core and optional sequences, the timing of imaging across disease phases, the use of imaging to separate true relapses from pseudo-relapses, and the particular diagnostic difficulty of seronegative neuromyelitis optica spectrum disorder, where the authors argue that new imaging biomarkers are still needed. Running alongside that, JAMA Neurology publishes a review from Flanagan and colleagues focused specifically on the MRI spectrum of MOGAD [2]. The emphasis there is on lesion dynamics, what they call radiologic lag, and the frequent resolution of T2 lesions, which stands in contrast to the persistent lesions typical of multiple sclerosis. Their most practice-relevant observation is that asymptomatic radiological activity is rare in MOGAD, which leads them to conclude that surveillance imaging has less clinical utility than it does in multiple sclerosis and less value as a trial surrogate. Taken together, these two documents are consensus and narrative synthesis rather than new data, but they are the clearest signal yet that the imaging strategy appropriate for multiple sclerosis should not simply be transplanted onto antibody-mediated disease.

Staying with imaging, Neurology reports a large cross-sectional multicentre study from Jain and colleagues that tries to separate ageing from disease in cervical cord atrophy [3]. Using scans from 480 healthy controls and nearly 1,300 patients with multiple sclerosis across three Italian sites, the investigators modelled the normal lifespan trajectory of upper cervical cord area, which rises until the late thirties and then declines after age 50, and then asked how far patients deviate from that curve. Patients with multiple sclerosis fell below the expected value from age 30 onward, with the gap most marked in the fifth and sixth decades before attenuating later as normal ageing catches up. The deviation did not differ between men and women, and although pediatric-onset patients looked worse at first glance, that difference disappeared once disease duration was accounted for. Greater cord atrophy tracked with higher disability and worse brain imaging measures, though the correlations were modest. The authors conclude that cord atrophy is a disease-specific marker of neurodegeneration beyond ageing, which supports its use as a biomarker, but this is cross-sectional work and not a longitudinal validation.

The second theme this week is precision diagnosis, and it runs from nosology through to single-gene discovery. Epilepsia publishes a report from the International League Against Epilepsy terminology commission, led by Specchio and colleagues, arguing that classification of the epilepsies should shift decisively toward etiology [4]. With more than a thousand epilepsy-related genes now identified, and with a single gene capable of producing several distinct phenotypes, the commission proposes extending the concept of etiology-specific epilepsy syndromes into what they call an etiology-syndromic spectrum model. They are candid about the obstacles, including the difficulty of pinning down genotype-phenotype relationships and the limited access to genetic and advanced imaging testing in resource-limited settings. This is a position paper, not a validated classification, but it signals where the formal taxonomy is heading.

Two papers in Movement Disorders show that shift in practice. Lüth and colleagues examined cerebrospinal fluid alpha-synuclein seed amplification assay results in carriers of the LRRK2 G2019S variant from a Norwegian cohort [5]. Among those tested, seeding was present in 80 percent of carriers who had Parkinson's disease and in only a single unaffected carrier, giving near-complete separation between the two groups. The investigators also report an exploratory association between a mitochondrial polygenic score and both the presence and the kinetics of seeding, replicated across two cohorts. The numbers in the seeding analysis are small, a few dozen participants in total, so this is hypothesis-generating, but it does suggest the assay can distinguish manifest from unaffected genetic carriers. The same journal reports a multisite imaging study from Santos and colleagues on spinocerebellar ataxia 27B, the recently described repeat expansion ataxia [6]. In 28 genetically confirmed patients compared with matched controls, the structural signature extended well beyond the cerebellum: there was volume loss in the cerebellar vermis, widespread reduction in white matter microstructural integrity in the subcortical cerebral white matter, and loss of grey matter area in the cervical spinal cord with relative preservation of cord white matter. It is a small cross-sectional cohort, but it argues that this is not a purely cerebellar disease.

And from Annals of Neurology, Fernández-Eulate and colleagues describe a genuinely new neuromuscular phenotype [7]. Across nine unrelated families, twelve patients with biallelic PIGB variants, a gene involved in glycosylphosphatidylinositol anchor biosynthesis, presented mostly with distal lower-limb weakness. Every patient had a motor-predominant neuropathy with conduction blocks on neurophysiology, nine showed signs of peripheral nerve hyperexcitability, and three had a decremental response on repetitive stimulation, two of whom were symptomatic and improved with pyridostigmine. Thirteen of the fifteen variants were novel, including a de novo inversion creating a gene fusion, and flow cytometry showed consistent reduction of free glycosylphosphatidylinositol in blood cells. The authors suggest that anchor biosynthesis genes belong in the differential for motor neuropathy with conduction blocks, which matters because that picture is often treated as acquired and immune-mediated.

The final theme covers prevention, prognosis and how we generate evidence. JAMA Neurology publishes the CHOICE randomized trial from Chen and colleagues, which asked whether fitting hearing aids slows cognitive decline [8]. Over 700 adults aged 60 and over in Shanghai, all with mild cognitive impairment and moderate to severe hearing loss, were randomized to hearing aids or to hearing care education and followed for 24 months. The primary outcome, conversion to dementia-level impairment, was not significantly different between groups, at roughly 3 percent with hearing aids and just under 5 percent with education alone. The trial was negative on its primary endpoint. What complicates the picture is a prespecified secondary finding: the proportion of participants whose cognitive rating improved to normal was about 15 percent in the hearing aid group against roughly 2 percent in the control group, a severalfold difference. That is a secondary outcome in an open-label trial with a low event rate for the primary endpoint, so it should be read as hypothesis-generating rather than as evidence that hearing aids prevent dementia.

Two papers address vascular neurology. In Neurology, Fandler-Höfler and colleagues pooled 82 patients with iatrogenic cerebral amyloid angiopathy from four European countries [9]. These patients presented at a median age of 50, a median of 39 years after presumed exposure, and over nearly 400 patient-years of follow-up the rate of symptomatic intracerebral haemorrhage was about 20 percent per year. Patients whose first presentation was haemorrhage had roughly triple the rate of recurrence compared with other presentations, while those whose exposure was spinal rather than cranial surgery had substantially lower haemorrhage rates. About half of patients had documented cognitive impairment at last assessment, though mortality was low at seven percent. This is retrospective pooled case-series data, but it is the best prognostic picture available for a condition that is being recognised more often. Finally, Stroke carries an expert consensus statement from Singh and colleagues on how to design prehospital stroke treatment trials in conventional emergency medical services settings, addressing patient identification, consent models, randomization, endpoint selection and statistical frameworks [10]. It is methodological guidance rather than clinical evidence, but it matters for anyone trying to move treatment earlier in the chain.

If you only have time for one paper this week, make it the CHOICE trial in JAMA Neurology [8]. It is the most direct randomized test yet of whether correcting hearing loss alters the trajectory from mild cognitive impairment to dementia, and its negative primary result reopens a question that observational epidemiology had made many clinicians consider settled.

Here is what this week's evidence adds up to in Neurology. First, hearing intervention in older adults with coexisting hearing loss and mild cognitive impairment did not significantly reduce conversion to dementia over two years in a single multicentre trial; a secondary signal of cognitive improvement is intriguing but unconfirmed. Second, expert consensus from two separate groups now argues that imaging strategy in MOGAD and neuromyelitis optica should diverge from the multiple sclerosis template, particularly around the limited yield of surveillance scanning in MOGAD, though this rests on expert opinion rather than trial data. Third, cervical cord atrophy in multiple sclerosis exceeds what ageing alone explains, most clearly in midlife, supporting its biomarker role, with longitudinal confirmation still outstanding. Fourth, genetics and seeding assays continue to redraw diagnostic boundaries, from the proposed etiology-based epilepsy classification, to alpha-synuclein seeding separating manifest from unaffected LRRK2 carriers in small numbers, to a new PIGB-related motor neuropathy with conduction blocks that could otherwise be mistaken for acquired disease. And fifth, iatrogenic cerebral amyloid angiopathy carries a high annual haemorrhage risk in retrospective pooled data, with risk concentrated among those presenting with haemorrhage and those exposed through cranial procedures.

That's your roundup for This Week in Neurology. The full transcript and references are available on the episode page in your AudioScholar library. This is an AI-curated summary — for clinical decisions, always consult primary sources and current guidelines. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And if it saves you time, a quick rating in your podcast app helps other physicians find the show.

This is an automated summary generated by artificial intelligence, which can make mistakes. Always review the original source materials.