This Week in Neurology — Sep 2, 2026

Welcome to This Week in Neurology. This week we're covering 10 notable papers spanning migraine prevention guidance, cerebrovascular risk and acute stroke therapy, and the practical management of neurodegenerative and immune-mediated disease. Let's dive in.

We start with the biggest piece of guidance to land this week. Neurology has published a paired guideline and systematic review from the American Academy of Neurology Guidelines Subcommittee and the American Headache Society on pharmacologic migraine prevention in adults [1] [2]. The systematic review, led by Pringsheim, pulled together 217 randomised trials searched through June of 2024, with outcomes centred on monthly headache days, the proportion of patients achieving at least a fifty percent reduction, and validated quality-of-life instruments [2]. For episodic migraine, the highest-confidence evidence sits with galcanezumab and erenumab, with moderate-confidence support for atogepant, eptinezumab, fremanezumab, and — importantly for cost and access — propranolol, topiramate and valproate. A long tail of older oral agents, including amitriptyline, metoprolol, bisoprolol, flunarizine, fluoxetine, levetiracetam, nifedipine, pizotifen and telmisartan, carried only low-confidence evidence of possible benefit. For chronic migraine, high-confidence evidence supported fremanezumab, galcanezumab and onabotulinumtoxinA, with moderate confidence for atogepant, eptinezumab, erenumab, topiramate and valproate. Across both populations, the monoclonals, the gepants, topiramate and onabotulinumtoxinA all improved patient-reported quality of life. The single most clinically consequential caveat is what the review could not answer: head-to-head comparisons between active treatments were sparse and of low or very low confidence, so there is still no evidence-based hierarchy telling you that a calcitonin gene-related peptide agent outperforms a well-tolerated beta-blocker in a given patient. The companion guideline, led by Potrebic, translates this into recommendations on when to start prevention, how to choose an agent, and — the part most likely to change a clinic afternoon — how to individualise in specific situations, including patients with fibromyalgia, obesity or hypertension, older adults, pregnancy and lactation, sex-related considerations, and patients with medication overuse, plus explicit guidance on assessing efficacy, monitoring adverse effects, and when to stop preventive treatment [1].

Turning to cerebrovascular disease, two papers address very different ends of the risk spectrum. In JAMA Neurology, the STRATEGY trial asked whether intensifying antiplatelet therapy can prevent early deterioration in branch atheromatous disease, a stroke subtype notorious for worsening in the first days [3]. Wang and colleagues randomised 970 patients across 38 hospitals in China, all with magnetic-resonance-confirmed branch atheromatous disease within 48 hours of onset, to intravenous tirofiban plus aspirin or placebo plus aspirin. The primary outcome of early neurological deterioration within seven days or new stroke by 90 days occurred in about 17 percent of the tirofiban group versus roughly 20 percent on aspirin alone — a difference that was not statistically significant. This was a negative trial. Tirofiban was reassuringly safe, with a single moderate or severe bleed in the whole treatment arm, but safety without efficacy is not a reason to adopt it. For now, aspirin remains the standard here, and the clinical problem of early deterioration in branch atheromatous disease remains unsolved. Meanwhile, The Lancet Neurology published a systematic review and meta-analysis from Dremel and colleagues on the prevalence of unruptured intracranial aneurysms, drawing on more than 316,000 participants and nearly 12,000 people with aneurysms [4]. Against a reference population of fifty-year-olds, prevalence was just under 4 percent. Risk was roughly four-fold higher in autosomal dominant polycystic kidney disease, where about one in eight patients harboured an aneurysm, and close to four-fold in connective-tissue disorders, at around one in ten. Current smoking, hypertension and female sex each raised risk more modestly. The striking finding is temporal: in healthy individuals imaged with magnetic resonance or computed tomographic angiography, prevalence in the most recent period studied was around 6.6 percent, nearly double the earlier period — and aneurysms of five millimetres or larger also became more common, which argues that better detection of tiny aneurysms and population ageing cannot fully explain the trend. Certainty of evidence ranged from very low to moderate with high heterogeneity, so treat the absolute numbers cautiously; the practical message is that the incidental aneurysm conversation is going to come up more often, and polycystic kidney disease, connective-tissue disease and family history remain your highest-yield screening triggers.

Our third theme is managing the neurodegenerative diseases we cannot yet stop. In Movement Disorders, Rubin and colleagues genotyped more than 2,000 patients with Parkinson's disease of Ashkenazi ancestry and linked them to national death registries [5]. Compared with idiopathic disease, glucocerebrosidase-mutation Parkinson's carried roughly a fifty percent higher mortality risk, while LRRK2-associated Parkinson's carried about a quarter lower risk — genuinely divergent prognoses within the same clinical diagnosis. Women had lower mortality. Median survival across all groups was 18 to 20 years from onset, and patients with early-onset disease lived longer with the disease but died younger, at a mean age of around 68 compared with 80 in late-onset disease. That is directly usable in prognostic conversations, and it makes genetic status relevant to counselling and not only to trial eligibility. On the inpatient side, Annals of Neurology reports what a dedicated hospital programme for people with Parkinson's disease can achieve [6]. Piccinin and colleagues combined an electronic health record census to identify admitted patients, monitoring and regimen alignment by movement-disorder-trained advanced practitioners, customised alerts and levodopa order sets, pharmacist support and staff education, then compared 366 admissions after implementation with a pre-implementation cohort. Improper levodopa formulation substitutions fell from nearly a fifth of cases to about five percent, timing deviations fell roughly twenty points to just over half, missed doses and dose deviations both declined, and among patients who received a contraindicated medication the median number of such doses dropped from two to one. Discharge to a non-home setting fell by about six percentage points. This was a before-and-after comparison against a historical cohort rather than a randomised trial, so secular change is a real limitation, but it is a concrete blueprint. Complementing this, Nature Reviews Neurology offers a synthesis from Ballard and colleagues on symptomatic treatment in Alzheimer disease, making the case that symptomatic care remains central even as disease-modifying therapies arrive [7]. Cognitive training, exercise, cognitive stimulation therapy and multimodal lifestyle programmes deliver small-to-moderate benefit, most convincingly in mild cognitive impairment. Cholinesterase inhibitors and memantine give modest but real benefit, and biomarker-stratified analyses suggest the gain is largest in those with confirmed Alzheimer pathology — an argument for letting biomarkers guide symptomatic prescribing too. For neuropsychiatric symptoms, personalised psychosocial interventions meaningfully reduce agitation and depression, while atypical antipsychotics offer only modest efficacy against substantial safety concerns; brexpiprazole for agitation and pimavanserin for psychosis are promising but need careful appraisal of effect size and safety.

Finally, three papers on refining the diagnosis. In Neurology, Liampas and colleagues pooled 20 studies and nearly 4,800 patients with multiple sclerosis to quantify slowly expanding lesions, the chronic active lesion marker visible on conventional magnetic resonance imaging [8]. These accounted for about 14 percent of all T2 lesions, but roughly four out of five patients had at least one, with a mean per-patient volume of about 1.4 millilitres against total T2 lesion volume of around 10 millilitres. Only about one in nine slowly expanding lesions overlapped with a paramagnetic rim lesion, so these two markers of chronic inflammation are related but far from interchangeable. Heterogeneity was high and identification methods varied, so this is a framework for interpreting the literature rather than a threshold for clinical use. Two teaching cases round out the week, both in Neurology. Martinez Sosa and colleagues describe a 57-year-old man with two months of unexplained gastrointestinal symptoms who then developed tremor, myoclonus, ataxia, hyperekplexia, insomnia, memory impairment and ocular flutter, with brainstem and limbic changes on imaging and a pleocytic spinal fluid — dipeptidyl-peptidase-like protein-6 antibody encephalitis, which responded dramatically to steroids, plasma exchange and rituximab [9]. Remember that gastrointestinal prodrome; it is the tell. And Di Pietro and colleagues show how nerve ultrasound and skin biopsy reframed a 74-year-old man's apparently idiopathic sensory axonal polyneuropathy, with reduced upper-limb nerve cross-sectional area and loss of somatic intraepidermal fibres with preserved autonomic innervation pointing towards targeted genetic testing [10].

If you only have time for one paper this week, make it the American Academy of Neurology and American Headache Society guideline on pharmacologic migraine prevention in adults [1]. Migraine is the highest-volume condition most neurologists treat, and this is the document that will define defensible prescribing, and defensible discontinuation, for the next several years.

Here are the key takeaways from this week in Neurology. First, the migraine prevention evidence base is now strongest for the calcitonin gene-related peptide agents and onabotulinumtoxinA, but with no reliable head-to-head data, propranolol, topiramate and valproate remain legitimate first choices. Second, intravenous tirofiban added to aspirin did not reduce early deterioration or recurrent stroke in branch atheromatous disease — do not adopt it. Third, unruptured aneurysm prevalence appears to be rising, and polycystic kidney disease, connective-tissue disorders and family history remain the highest-yield reasons to look. Fourth, genetic status carries prognostic weight in Parkinson's disease, with glucocerebrosidase mutations conferring higher and LRRK2 mutations lower mortality than idiopathic disease. And fifth, a structured inpatient Parkinson's programme measurably reduced levodopa errors and discharges to non-home settings — a systems intervention worth taking to your hospital leadership.

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.