PILLAR
Prospective Trial of Cerebrospinal Fluid Filtration after Aneurysmal Subarachnoid Hemorrhage via Lumbar Catheter (PILLAR)
Year of Publication: 2019
Authors: Spiros L. Blackburn, Andrew W. Grande, Christa B. Swisher, ..., J. Javier Provencio
Journal: Stroke
Citation: Stroke. 2019 September; 50(9): 2558-2561. doi:10.1161/STROKEAHA.119.025399.
Link: https://pmc.ncbi.nlm.nih.gov/articles/PMC6710124/
PDF: https://www.ahajournals.org/doi/pdf/10.1...EAHA.119.025399
PDF: Clinical Question
Is Neurapheresis CSF filtration via a dual-lumen lumbar catheter safe, tolerable, and feasible for removing blood and blood breakdown products from CSF after aneurysmal subarachnoid hemorrhage?
Study Overview
Objective
To evaluate the safety, tolerability, and feasibility of Neurapheresis cerebrospinal fluid (CSF) filtration therapy to remove blood and blood byproducts in patients with aneurysmal subarachnoid hemorrhage (aSAH).
Study Summary
- Neurapheresis was feasible and well-tolerated in aSAH patients, with significant reductions in CSF red blood cell count and Hijdra score indicating blood clearance.
- No serious adverse events related to the device were observed.
Intervention
Insertion of a dual-lumen lumbar catheter following aneurysm securement to perform Neurapheresis CSF filtration for up to 24 hours. Filtered CSF was returned to the thoracic subarachnoid space.
Patients per Arm
13 patients received Neurapheresis therapy (out of 15 consented).
Bottom Line
The PILLAR study, a first-in-human feasibility trial, demonstrated that Neurapheresis CSF filtration via a dual-lumen lumbar catheter is safe and feasible in aneurysmal subarachnoid hemorrhage. In 13 treated patients (of 15 consented), mean CSF RBC count fell from 2.78×10⁵ to 1.17×10⁵ cells/µL (52.9% reduction), total protein was reduced 71%, and mean Hijdra CT score dropped from 22 to 12 (46.5% mean reduction), with only 3 minor anticipated adverse events and no device-related serious AEs or UADEs.
Major Points
- 15 patients with aneurysmal SAH were consented; 13 received Neurapheresis therapy (1 catheter placement not attempted, 1 placement unsuccessful) and were followed for 30 days.
- Primary endpoint was safety and feasibility (AEs associated with catheter placement/removal and operation of the Neurapheresis system); the paper explicitly states no formal statistical analysis was performed due to lack of controls and small enrollment.
- No device-related serious AEs and no unanticipated adverse device effects (UADEs) occurred; the paper explicitly states no AEs were associated with catheter placement. 3 minor, anticipated AEs were reported (one back/leg pain noted at 2-week follow-up and resolved by 30 days; one transient vomiting with headache and unremarkable head CT; one transient vomiting after CSF sampling).
- Mean catheter placement was 24:13 h after aneurysm rupture; catheter remained in situ a mean of 30:23 h; aggregate Neurapheresis pump time averaged 15:07 h (range 5:32–24:00 h); only 1 subject reached the maximum 24 h of pump time.
- Mean CSF processed 632.0 mL (median 530.0, range 180.6–1447.6 mL), with 439.0 mL returned to the patient and mean 193.0 mL (range 75.8–328.1 mL) of concentrated waste removed.
- Mean CSF RBC count fell from 2.78×10⁵ to 1.17×10⁵ cells/µL (52.9% reduction); total protein was reduced 71%; 53.8% of patients had total protein ≤100 mg/dL at end of therapy; CSF glucose remained normal.
- In the 3 patients with >22 h of filtration, mean RBC reduction was 79.7% (complete RBC clearance not achieved).
- Independent CT analysis showed mean Hijdra score dropped from 22 (out of 30) to 12 (46.5% mean reduction).
- Neurological (GCS, WFNS, Hunt & Hess) and peripheral checks remained stable or improved in 92.3% (12/13) of patients during therapy.
- At 30 days: 61.5% (8/13) had GOS 5; 69.2% (9/13) had mRS 0–2 and Barthel Index >90; mean ICU stay 12 days (range 6–18).
Design
Study Type: Prospective, industry-sponsored, single-arm, multicenter, first-in-human feasibility trial (open-label, non-randomized)
Randomization:
Enrollment Period: Not specified in the paper (published September 2019)
Follow-up Duration: Clinical follow-up at 2 weeks (±3 days) and 30 days (±7 days) post-catheter removal; catheter indwelling mean 30:23 h; aggregate pump time mean 15:07 h (range 5:32–24:00).
Centers: 4
Countries: United States
Sample Size: 15 consented; 13 received Neurapheresis therapy (1 catheter placement not attempted, 1 unsuccessful)
Analysis: No formal statistical analysis was performed; the paper explicitly states: 'A statistical analysis was not performed due to lack of controls and small enrollment number.' Baseline and post-Neurapheresis CT images were scored by an independent neuroradiologist using the Hijdra scale; descriptive statistics only.
Inclusion Criteria
- Diagnosis of aneurysmal subarachnoid hemorrhage (aSAH).
- Ruptured aneurysm secured prior to enrollment (allowing lumbar catheter placement).
- Informed consent from patient or legally authorized representative.
- (Full criteria are provided in the paper's Supplemental Table I, not reproduced in the main text.)
Exclusion Criteria
- Full exclusion criteria are provided in the paper's Supplemental Table I; the main text does not list them.
Baseline Characteristics
| Characteristic | Control | Active |
|---|---|---|
| CSF RBC count, mean (initial), cells/µL | 2.78×10⁵ | |
| CSF Hijdra sum score, mean (initial, out of 30) | 22 | |
| Time from aneurysm rupture to catheter placement, mean, h:mm | 24:13 |
Arms
| Field | CSF Filtration (Neurapheresis System) |
|---|---|
| Intervention | Dual-lumen intrathecal lumbar catheter placed under fluoroscopy after aneurysm securement, connected to an extracorporeal Neurapheresis filtration unit (electronic valve, flow meter, filter, pressure sensors, pump); aspirated blood-contaminated CSF from the lumbar cistern and returned filtered CSF to the thoracic subarachnoid space; waste rate typically 12–15 mL/h; protocol limit 24 h aggregate pump time or 36 h catheter indwelling. |
| Duration | Catheter indwelling mean 30:23 h; aggregate Neurapheresis pump time mean 15:07 h (range 5:32–24:00 h). |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Safety, tolerability, and feasibility of Neurapheresis CSF filtration therapy via lumbar catheter (up to 24 h of aggregate pump time) in aneurysmal subarachnoid hemorrhage, based on adverse events associated with catheter placement/removal and system operation, and any unanticipated adverse device effects (UADEs). | Primary | 15 consented, 13 (86.7%) had a catheter placed and underwent Neurapheresis therapy (1 placement not attempted, 1 unsuccessful); mean pump time 15:07 h (range 5:32–24:00); mean 632.0 mL CSF processed (range 180.6–1447.6); 3 minor, anticipated AEs, no device-related serious AEs or UADEs; neurologic/peripheral checks stable or improved in 92.3% (12/13); mean Hijdra score 22 → 12 (46.5% reduction); CSF RBC 2.78×10⁵ → 1.17×10⁵ cells/µL (52.9% reduction); total protein reduced 71%. | No statistical analysis was performed (paper explicitly states this due to lack of controls and small enrollment). | ||
| Change in mean CSF red blood cell count | Secondary | 2.78×10⁵ → 1.17×10⁵ cells/µL (52.9% reduction) | Not reported (no statistical analysis performed) | ||
| Change in CSF total protein | Secondary | 71% reduction; 53.8% of patients had total protein ≤100 mg/dL at end of therapy | Not reported | ||
| Change in Hijdra CT score (independent neuroradiologist scoring) | Secondary | Mean 22 (out of 30) → 12 (46.5% mean reduction) | Not reported | ||
| 30-day functional outcome | Secondary | GOS 5 in 61.5% (8/13); mRS 0–2 and Barthel Index >90 in 69.2% (9/13); mean ICU stay 12 days (range 6–18) | |||
| Any adverse event | Adverse | 3/13 (23.1%) — all single events, minor, anticipated; paper states no AEs were associated with catheter placement | |||
| Back and leg pain | Adverse | 1/13 (7.7%) — noted at 2-week follow-up, resolved by 30-day follow-up | |||
| Transient vomiting with concurrent headache | Adverse | 1/13 (7.7%) — head CT unremarkable | |||
| Transient vomiting after CSF sampling | Adverse | 1/13 (7.7%) | |||
| Device-related serious AEs / UADEs | Adverse | 0/13 (0%) | |||
| Premature termination (not classified as an AE by DSMB) | Adverse | 1 patient had Neurapheresis terminated early due to neurologic decline, focal cerebral edema, and midline shift attributed to the surgical clipping procedure |
Criticisms
- Small, non-randomized, non-controlled, single-arm feasibility study (n=13 treated); cannot establish efficacy or make comparative conclusions.
- Paper explicitly states no formal statistical analysis was performed due to lack of controls and small enrollment.
- Pump time was highly variable — only 1 subject received the full 24 h of protocol therapy (pump was stopped for standard-of-care activities and by the 36 h catheter-indwelling cap).
- Focused on safety, tolerability, and biochemical/radiographic CSF endpoints; not powered for clinical outcomes such as delayed cerebral ischemia, vasospasm, or long-term functional outcome.
- Industry-sponsored (Minnetronix Neuro, Inc.); all authors received support for approved study-related costs.
- Complete RBC clearance was not achieved even in patients with >22 h of filtration (mean RBC reduction 79.7%), motivating the subsequent PILLAR-XT study (NCT03607825).
Subgroup Analysis
None. In 3 patients with >22 h of filtration, mean RBC reduction was 79.7% (complete RBC clearance not achieved).
Funding
Industry-sponsored by Minnetronix Neuro, Inc. Dr. Blackburn's effort supported by NIH K23NS106054.
Based on: PILLAR (Stroke, 2019)
Authors: Spiros L. Blackburn, Andrew W. Grande, Christa B. Swisher, ..., J. Javier Provencio
Citation: Stroke. 2019 September; 50(9): 2558-2561. doi:10.1161/STROKEAHA.119.025399.
Content summarized and formatted by NeuroTrials.ai.