What we are reading

One thousand patients with acute back pain at high risk of becoming chronic: supported self-management lowers disability at one year and spinal manipulation alone does not. Pain intensity moved in none of the four arms, and that is worth saying before we refer.

Spinal Manipulation and Clinician-Supported Biopsychosocial Self-Management for Acute Back Pain: The PACBACK Randomized Clinical Trial

JAMA · December 2025

Irradiating the cavity during the same operation added nothing: progression-free survival was unchanged and recurrence remained local in three of every four patients, with more radiation necrosis and more severe seizures. The phase 3 halts a line that phase 1/2 had left looking promising.

Dose escalation with intraoperative radiotherapy in newly diagnosed glioblastoma (INTRAGO-II): an open-label, multicentre, randomised, controlled, phase 3 trial

The Lancet Oncology · June 2026 · open access

Funded in part by Carl Zeiss Meditec, manufacturer of the intraoperative radiotherapy device under test. The result runs against the sponsor.

Earlier blood-pressure trials after thrombectomy failed because they lowered pressure the same way for everyone. Setting the target by degree of reperfusion took 90-day independence from 47% to 60% and reduced haemorrhagic transformation, with no excess symptomatic bleeding or mortality. It costs nothing to apply.

Personalized Blood Pressure Targeting After Endovascular Therapy for Acute Ischemic Stroke: A Randomized Clinical Trial

JAMA Neurology · June 2026

The first prospective international comparison of the four temporising procedures actually used in the preterm infant with intraventricular haemorrhage. Across 238 neonates the difference is not in complications but in eventual shunt dependence: neuroendoscopic lavage left the fewest shunts at six months.

Treatment of Posthemorrhagic Hydrocephalus in neonates (TROPHY) registry: surgical results from 6-month follow-up data

Journal of Neurosurgery: Pediatrics · December 2025

Across 758 patients the robot does not place electrodes more accurately than the stereotactic frame: target, depth and radial errors are indistinguishable, and so are complications. What it buys is 33 minutes per case. A unit without a robot does not implant worse, it implants more slowly.

Robot-assisted versus frame-based stereoelectroencephalography (sEEG) electrode implantation in drug-resistant epilepsy: a meta-analysis of accuracy, efficiency, and safety

Acta Neurochirurgica · February 2026 · open access

Faced with a large acute epidural haematoma with herniation, leaving the bone out is almost a reflex. The trial finds no functional advantage at six months over craniotomy with flap replacement, and nearly three times more delayed intracranial haemorrhage. With 120 patients it does not prove equivalence, but it shifts the burden of proof.

Safety and efficacy of decompressive craniectomy versus standard craniotomy for large acute epidural haematoma with tentorial herniation in China (PREDICT-AEDH): a nationwide, multicentre, open-label, parallel-group, randomised controlled trial

The Lancet Neurology · June 2026