This week in Neurosurgery

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

This week: leg compression before induction cuts hypotension during trauma craniotomy, a day-one blood marker tracks later intracranial pressure, and gliomas labelled inoperable elsewhere were resected awake with comparable outcomes.

Welcome to This Week in Neurosurgery. We're covering 10 papers this week: 3 in depth, 4 in brief, 3 quick signals.

In the Journal of Clinical Neuroscience, Ahlawat and colleagues report a randomised, double-blind trial of pneumatic leg compression in adults aged 18 to 65 with isolated severe traumatic brain injury undergoing decompressive craniotomy [1]. Ninety patients were randomised, and 89 analysed, to a compression device switched on before induction or an identical device left inactive. Post-induction hypotension, defined as a systolic pressure below 90 within 20 minutes of induction, occurred in about a third of patients with active compression compared with just under 60 percent of controls, a risk reduction of a little under half. Fewer patients in the compression group needed mephentermine or phenylephrine, and no device-related complications were seen. Baseline systolic pressure was already higher in the active group after ten minutes of compression, but adjusting for it did not change the result. Hypotension after the first 20 minutes, lactate and acid-base measures did not differ. The main limitation is that every outcome is haemodynamic: the trial did not measure cerebral perfusion or neurological recovery, and the sample is modest. It is the first randomised evidence for this simple measure in neurotrauma, and the authors call for larger multicentre trials.

Staying with traumatic brain injury, Bonow and colleagues, writing in Neurosurgery, used the Transforming Research and Clinical Knowledge in Traumatic Brain Injury data set for a retrospective cohort of 229 patients aged 17 and older with invasive intracranial pressure monitoring, excluding those with craniectomy on admission [2]. They asked whether day-one blood levels of glial fibrillary acidic protein, known as GFAP, and of UCH-L1 tracked later intracranial pressure and treatment intensity. Patients in the highest GFAP quartile spent about 23 percent of readings above 22 millimetres of mercury, compared with about 7 percent in the lowest quartile. Secondary decompressive craniectomy was performed in 39 percent of the highest quartile against 16 percent of the lowest. After adjustment, GFAP remained independently associated with intracranial pressure burden and with therapeutic intensity over five days. UCH-L1 showed no significant relationship. The limitations are the retrospective design, a modest monitored sample, and no tested threshold or prospective decision rule. The authors frame GFAP as a possible adjunct for risk stratification, and the evidence so far is association, not biomarker-guided management.

Elia and colleagues, in the Journal of Neuro-Oncology, reviewed 448 adults who underwent awake function-based craniotomy for diffuse glioma between 2011 and 2024 [3]. Sixty-six patients, about 15 percent, had been told by an outside institution that their tumour was not operable. These patients arrived with more prior oncological treatment, more focal deficits, larger tumours, more midline shift and more IDH-wildtype glioblastoma. Despite that, intraoperative adverse events, surgical complications, six-month Karnofsky score, seizure control, extent of resection and residual volume did not differ significantly from patients considered operable from the outset, and propensity matching gave the same result. No awake procedure was aborted in the previously inoperable group. Voxel-based lesion mapping found no anatomical location that independently predicted the inoperable label, which the authors read as evidence that the designation partly reflects centre judgment. The limitations are a retrospective design in a specialised high-volume setting and a selected group of patients who sought a second opinion and were accepted for surgery. It adds evidence that operability in diffuse glioma depends partly on where a patient is assessed.

Now, four in brief.

Returning to traumatic brain injury, Mohan and colleagues in Acta Neurochirurgica pooled six cohorts with 6,825 adults, all with a Glasgow Coma Scale of 8 or less, comparing early with late pharmacological thromboprophylaxis [4]. Earlier prophylaxis was associated with roughly half the odds of venous thromboembolism, but at borderline significance that disappeared when either large cohort was removed or when the effect was expressed as a risk ratio. In-hospital mortality did not differ. Data on haemorrhage progression came from a single cohort with three events. Timing definitions varied between studies and certainty was very low for every outcome, so the safety question remains open pending randomised trials.

Two papers this week address radiation for resected brain metastases. Kite and colleagues, in the Journal of Neuro-Oncology, meta-analysed fifteen pre-operative radiosurgery studies and four cesium-131 brachytherapy studies, about eleven hundred patients in total [5]. Twelve-month local control was about 97 percent with brachytherapy and 90 percent with pre-operative radiosurgery, a difference that reached significance only in some analyses. Hypofractionated pre-operative radiosurgery matched brachytherapy while single-fraction treatment lagged, suggesting dose rather than modality drives the gap. Leptomeningeal disease and radiation necrosis did not differ significantly. The comparison is indirect, across separate sets of studies, and the review was not registered.

Also in the Journal of Neuro-Oncology, Sinha and colleagues reviewed 182 patients receiving cavity-directed Gamma Knife radiosurgery after resection [6]. Treatment within 14 days was associated with median survival of about 38 months against about 15 months with later treatment, and this held after multivariable adjustment. Local recurrence-free survival was not associated with timing, and the 14-day threshold lost significance once corrected for the data-driven way it was chosen. A survival difference without a local-control difference is hard to explain mechanistically, so the finding is hypothesis-generating.

In the Journal of Neurosurgery, Hallan and colleagues analysed the NeuroVascular Quality Initiative-Quality Outcomes Database registry for 401 unruptured aneurysms treated with intrasaccular flow disruption, 72 of them in off-label locations [7]. Good discharge function was seen in over 94 percent of both groups, intraoperative complications were around 3 percent, and complete occlusion at discharge was a little over half in each group, with no significant differences, confirmed by propensity matching. Missing data prevented analysis of occlusion at last follow-up, so the registry supports short-term safety in off-label locations while durability remains unknown.

Three quick signals.

In Neurosurgery, Ma and colleagues described 80 large core thrombectomies from one prospectively maintained database [8]. Patients older than 75 and treated more than 12 hours after last known well had no independent walkers at 90 days, and 62 percent of those patients died. With only 80 patients, these risk categories are descriptive and need validation.

Pääkkönen and colleagues, also in Neurosurgery, followed shunted patients with idiopathic normal pressure hydrocephalus in a prospective population-based cohort [9]. The Gothenburg scale was the most responsive measure and discriminated gait improvement well at 12 months, while the 15D quality-of-life instrument missed disease-specific change. The abstract does not report the cohort size.

And Alani and colleagues, in Neurosurgery, found that among 46 patients repaired for spontaneous cerebrospinal fluid leaks, opening pressure rose by about 5 centimetres of water on average after closure [10]. Every patient with a rise of a quarter or more had an active leak at surgery. This is a small retrospective series from one institution.

Stepping back across this week's papers, most of the evidence rests on retrospective cohorts, registries and pooled observational data, and the one randomised trial measured blood pressure and drug use rather than brain outcomes. Several findings depend on thresholds chosen after looking at the data or on comparisons across separate study sets. Where sensitivity analyses were reported, signals often weakened, which leaves open questions about timing, dosing and patient selection that only prospective comparative work can close.

If you only have time for one paper this week, make it the awake resection study from Elia and colleagues [3]. It reopens the question of whether a non-operable label in diffuse glioma reflects the tumour or the centre making the call.

That's your roundup for This Week in Neurosurgery. The full transcript and references are available on the episode page. See you next week.

If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And a quick rating on Apple Podcasts or Spotify helps other physicians discover the show.

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