← Back
NeuroTrials.ai
Neurology Clinical Trial Database

PILLAR-XT

Prospective Trial of Cerebrospinal Fluid Filtration After Aneurysmal Subarachnoid Hemorrhage via Lumbar Catheter Extension

Share Share Copied! Compare

Year of Publication: 2025

Authors: Spiros L. Blackburn, Marc A. Babi, Andrew W. Grande, ..., Aaron R. McCabe

Journal: Neurocritical Care

Citation: Neurocrit Care. 2025. https://doi.org/10.1007/s12028-025-02328-8

Link: https://doi.org/10.1007/s12028-025-02328-8

Clinical Question

What is the safety, tolerability, and filtration capability of the Neurapheresis CSF Management System for removing blood and blood lysis products from hemorrhagic CSF in patients with aneurysmal subarachnoid hemorrhage? (The study did not evaluate impact on clinical outcome.)

Bottom Line

PILLAR-XT, a prospective single-arm IDE study, showed that extracorporeal CSF filtration with the Neurapheresis system was safe and feasible in aSAH, achieving large within-cohort reductions in CSF red blood cells (86%), CSF protein (82%), and Hijdra Sum Score (65%) from first to last measurement. The study did not evaluate clinical outcomes and did not directly compare CSF blood/protein levels with EVD or lumbar drain controls.

Major Points

  • PILLAR-XT was a prospective, single-arm, investigational device exemption study at 6 US sites (NCT03607825)
  • 33 patients enrolled; 29 had catheter placement attempted; 27 successfully treated (93% technical success)
  • Dual-lumen intrathecal catheter placed at L3-L4 or L4-L5 with extracorporeal filtration for up to 72 h
  • Study evaluated safety, tolerability, and filtration curve of blood and its lysis products; no primary efficacy endpoint was formally designated
  • Mean within-cohort reduction in Hijdra Sum Score from enrollment to catheter removal was 65% (25.6 to 7.5 median; p<0.001)
  • Mean within-cohort reductions in CSF RBC and protein from first to last measurement were 86% and 82%, respectively (p=0.003)
  • Median filtration time 37 h (IQR 24:03-38:52) with median waste rate 5.7 mL/hr
  • Compared with published standard-of-care data (Koopman et al.), Neurapheresis-treated patients reached normal median CSF protein at ~2.1 days after ictus, whereas the standard-of-care cohort did not reach normal within 20 days (qualitative external comparison, not an in-trial control)
  • 5 adverse events in 4 patients (1 SAE = transtentorial herniation, resolved); all mild-moderate with no clinical sequelae
  • 92% of patients maintained or improved GCS from admission to ICU discharge; 64% were at home at 30 days
  • The study did not evaluate the impact of CSF filtration on clinical outcome, and did not directly compare CSF RBC/protein levels with EVD or lumbar-drain patients

Design

Study Type: Prospective, single-arm, investigational device exemption study

Randomization:

Blinding: Open-label, no blinding

Enrollment Period: December 2018 - January 2021

Follow-up Duration: 30 days

Centers: 6

Countries: United States

Sample Size: 27

Analysis: Paired t-tests for HSS and CSF RBC/protein (mean of pairwise reductions), descriptive statistics, Spearman rank correlation for HSS interrater reliability

Inclusion Criteria

  • Age 18-70 years
  • Informed consent by patient or legally authorized representative
  • Modified Fisher Grade 2, 3, or 4
  • Hunt & Hess I-IV
  • First aneurysmal SAH confirmed by CT scan and secured or planned securement via clipping or coiling per institutional SOC
  • Patient ≤48 hours post bleeding event
  • World Federation of Neurosurgical Societies (WFNS) Grades I-IV
  • Patient indicated for ventriculostomy

Exclusion Criteria

  • Coagulopathy that cannot be reversed per investigator discretion
  • Pregnancy
  • SAH due to mycotic aneurysm or arteriovenous malformation
  • Acute myocardial infarction or unstable angina
  • Uncontrolled diabetes at time of catheter placement
  • Creatinine >2.0 mg/dl
  • Supratentorial mass lesions >50 cc (early cohort) or ≥15 cc (later cohort)
  • More than 5 mm midline shift with infarction/edema (early cohort) or ≥2 mm midline shift with infarction/edema (later cohort)
  • Infratentorial mass lesion ≥10 cc
  • Presence of any subdural hematoma
  • Effacement of basilar cisterns (suprasellar, ambient, chiasmatic, quadrigeminal)
  • Vasospasm on admission by angiographic evidence
  • Connective tissue disorder that may impact dural integrity
  • Thrombocytopenia (platelet count <100,000)
  • Low molecular weight heparin (e.g., Lovenox)
  • Clopidogrel (Plavix) or other chronic platelet inhibitors
  • Documented history of cirrhosis
  • Non-communicating obstructive hydrocephalus
  • Lumbar/thoracic spinal anatomy (e.g., severe spinal stenosis) or posterior fusion hardware interfering with catheter placement
  • Existing hardware that prevents accurate CT imaging
  • Pre-existing lumbar drain
  • Local skin infections or eruptions over the puncture site
  • Signs of CNS systemic infection, sepsis, or pneumonia
  • Lumbar puncture within 6 h
  • Concurrent participation in another (non-observational/non-retrospective) study
  • Without prior approval from the Sponsor

Baseline Characteristics

CharacteristicControlActive
Total Enrolled33 patients
Successfully Treated27 patients
Mean Age49.9 ± 11.8 years
Female76% (22/29)
Currently Smoking48% (14/29)
Hypertension62% (18/29)
Diabetes17% (5/29)
Aneurysm Clipped38% (11/29)
Aneurysm Coiled62% (18/29)
Median WFNS Grade2
Modified Fisher Grade 324 patients
Modified Fisher Grade 45 patients
Hunt & Hess Grade 320 patients
Hunt & Hess Grade 43 patients

Arms

FieldNeurapheresis Treatment
InterventionNeurapheresis CSF Management System with a dual-lumen intrathecal catheter placed at L3-L4 or L4-L5; extracorporeal filtration aspirates lumbar CSF, filters out blood and lysis products, and returns filtered CSF to the thoracic subarachnoid space
DurationUp to 72 hours of filtration

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Safety, tolerability, and filtration curve of blood and its lysis products from hemorrhagic CSF (no single primary efficacy endpoint was formally designated)PrimarySafe and tolerable: 5 AEs in 4 patients (1 SAE, all resolved without sequelae); filtration achieved with mean 86% CSF RBC reduction, 82% CSF protein reduction, and 65% HSS reduction
Hijdra Sum Score reduction (enrollment to catheter removal)SecondaryMean 65% reduction (median 25.6 to 7.5)<0.001
CSF red blood cell reduction (first to last measurement)SecondaryMean 86% reduction (540×10⁹ to 30×10⁹ cells/L)0.003
CSF protein reduction (first to last measurement)SecondaryMean 82% reduction (3.99 to 0.56 g/L)0.003
Catheter placement successSecondary93% (27/29 attempted)
Median filtration timeSecondary37:00 h (IQR 24:03-38:52)
Patients maintaining/improving GCS (admission to ICU discharge)Secondary92% (25/27)
Patients at home at 30 daysSecondary64% (16/25)
Median Barthel Index at 30 daysSecondary95 (IQR 60-100)
Total adverse eventsAdverse5 events in 4 patients
Serious adverse eventsAdverse1 event (transtentorial herniation, resolved with standard clinical protocol)
Localized pain (legs/back)Adverse2 events
HeadachesAdverse2 events
New focal neurological disorderAdverse11% (3/27)
New cerebral infarct on imagingAdverse7% (2/27)
Delayed cerebral ischemiaAdverse11% (3/27)
EdemaAdverse11% (3/27)
SeizureAdverse4% (1/27)
CNS infection (within 5 days post catheter removal)Adverse0% (0/27)
Shunt-dependent hydrocephalusAdverse19% (5/27)

Subgroup Analysis

No prespecified subgroup analysis was reported. Related cohort-level observations: 20 of 26 evaluable patients returned to the normal CSF protein range (0.15-0.7 g/L) within 5.5 days of SAH onset, and all adverse events were mild-moderate with complete resolution.

Criticisms

  • Single-arm design without a randomized concurrent control group
  • Small sample size (27 patients successfully treated)
  • No direct comparison to standard lumbar drain or EVD RBC/protein levels; comparisons rely on previously published cohorts (e.g., Koopman et al.)
  • EVDs were opened in 11/27 patients during filtration (protocol allowed if ICP >20 mm Hg or per investigator judgment), which may confound clearance measures
  • Limited follow-up duration (30 days)
  • Study explicitly did not evaluate impact on clinical outcomes
  • Selection bias inherent in an IDE study; inclusion/exclusion criteria were adjusted mid-trial after the SAE

Funding

Supported by Minnetronix Neuro, Inc. and the National Institute of Neurological Disorders and Stroke (SBIR Grant R44NS110247)

Based on: PILLAR-XT (Neurocritical Care, 2025)

Authors: Spiros L. Blackburn, Marc A. Babi, Andrew W. Grande, ..., Aaron R. McCabe

Citation: Neurocrit Care. 2025. https://doi.org/10.1007/s12028-025-02328-8

Content summarized and formatted by NeuroTrials.ai.