Effectiveness of Lumbar Cerebrospinal Fluid Drain Among Patients With Aneurysmal Subarachnoid Hemorrhage

Wolf. JAMA Neurol. 2023; 80:833-842. doi:10.1001/jamaneurol.2023.1792

Clinical Question

• In patients with aneurysmal subarachnoid haemorrhage (aSAH), does lumbar CSF drainage compared to standard therapy improve clinical outcomes as measured by the modified Rankin Scale (mRS) score at 6 months?

Background

• aSAH frequently leads to death or permanent disability due to both the initial insult and delayed cerebral ischaemia (DCI)
• DCI occurs due to vasospasm in up to 70% of patients with SAH
• This may be related to blood degradation products in the CSF near the basal cisterns
• It is hypothesised that CSF drainage via a lumbar drain may preferentially remove more of the blood degradation products due the effect of gravity and thus may decrease rates of vasospasm and DCI
• Retrospective data and meta-analyses of lumbar drain use in aSAH has demonstrated lower mortality, and reduced rates of clinical vasospasm and ischaemic stroke and improved neurological outcomes
• LUMAS, a prospective single centre RCT, including patients with aSAH (WFNS grade 1-3), demonstrated improved rates of delayed ischaemia and early neurological outcome but no difference in outcome at 6 months

Design

• Parallel group, multi-centre, open label randomized clinical trial
• Randomization – 1:1 to standard care or additional use of lumbar drain
  • Via internet randomizer – permutated blocks of 6
  • No stratification
• Blinded end-point evaluation
• Patients followed up to their death or 6 months post randomisation
• Intention to treat analysis
• Protocol and analysis plan published prior
• Power calculation:
  • 300 patients would be required to detect a decrease in unfavourable outcome from 50 to 33% with 85% power

Setting

• 19 hospitals in Germany, Switzerland and Canada
  • All provided 24hr acute neurosurgical and neurocritical care with > 30 aneurysm procedures / year
• Data collected from Jan 2011 to Jan 2016

Population

• Inclusion:
  • ≥ 18yo, acute SAH with confirmed aneurysm origin via CT-angiography or DSA
  • Aneurysm treatment within 48hrs of SAH
• Exclusion
  • Presence of therapeutic anticoagulation
  • Contraindications for lumbar drain
  • Absent or compressed basal cisterns on admission CT-B
  • Pregnancy
  • Reduced life expectancy
  • Participation in another interventional trial
  • Haemorrhage from non-aneurysmal origin
  • No haemorrhage present on initial CT-B (Fischer 1/ modified Fischer grade 0)
• 307 patients randomised
  • Lumbar drain > 155 randomised  (11 excluded)
  • Standard care > 152 randomised (9 excluded)
• Baseline characteristics of lumbar drain vs. standard care
  • Similar between groups
  • Median age: 54 vs 56 years
  • Female: 68% vs 69%
  • mRS score of 0 on admission: 94 vs 93%
  • Presence of intraventricular haemorrhage (IVH):63% vs 59%
  • Presence of intracerebral haemorrhage (ICH): 39% vs 35%
  • WFNS classification:
    • 1 = 37% vs 29%
    • 2 = 15% vs 15%
    • 3 = 5 % vs 7%
    • 4 = 10% vs 11%
    • 5 = 33% vs 39%
  • Modified Fisher score of 4: 59 vs 55%
  • Location, size and number of aneurysms similar

Intervention

• Insertion of lumbar drain with drainage of lumbar CSF at 5ml/hr
  • Commenced within 72hrs of aneurysmal SAH in addition to standard therapy
  • Continued for least 4 days and up to 8 days
  • Safety: If ICP pressure >20mmHg or >5mmHg difference between ICP + lumbar drain – recommended to cease lumbar drainage
• Median CSF lumbar drainage of 108 mls in first 8 days

Control

• No lumbar drainage

Management common to both groups

• Treatment of aneurysmal SAH with endovascular coiling or clipping determined by local treating teams
• Local teams determined if EVD or ICPm inserted
  • EVDs inserted in 70.8% of lumbar drain group and 76.9% of control group
  • Median daily drainage via EVD lower in lumbar drain group (98 vs 171 mls)
  • Total CSF drainage via both EVD +/- lumbar drains similar between groups (eFigure 13)
• Local standard management for aSAH:
  • Daily transcranial doppler monitoring as per local protocol
  • Cerebral vascular imaging via CTA, MRA or DSA for suspected vasospasm or routine day 7 to 10
  • Treatment for vasospasm at discretion of local investigators – not protocolised
• Rates of aneurysm treatment (53 vs 50% coiled) and vasospasm prophylaxis (99 vs 99% nimodipine) use similar between groups
• MAP, fluid intake and fever burden similar between groups

Outcome

• Primary outcome (lumbar drain vs control):
  • Rate of unfavourable neurological outcome measured at 6 months (defined as mRS 3 – 6)
  • Statistically significant lower rate of unfavourable outcome
    • 32.6% vs 44.8%
    • RR 0.73 (95% CI: 0.51 – 0.98; p = 0.04)
  • Adjustment for age, severity via Hunt-Hess grade, and presence of ICH and IVH:
    • RR 0.76 (95% CI 0.54 – 1.00, p = 0.047)
  • Effect consistent across pre-specified subgroups (age, sex, grade, Location, IVH / ICH, treatments provided, centre size)
• Secondary outcomes (lumbar drain vs control):
  • Significant difference in:
    • Rate of secondary infarction following discharge
      • 28.5% vs 39.9% (RR 0.71, 95% CI 0.49 to 0.99, p = 0.04)
    • GOS-E 1-4 at 6 months and Barthel Index < 80 at 6 months
  • No significant difference in:
    • Mortality within 6 months
      • 13.2 vs 17.5%
    • Rate of vasospasm
      • Clinical: 29 vs 33%
      • TCD: 27 vs 25%
      • Angiographic: 46 vs 44%
    • Infection of any cause
      • 39 vs 36%

• Adverse outcomes
  • Lumbar drain group:
    • 1 patient developed elevated gradient >5mmHg between ICP + lumbar drain prohibiting drainage
    • 1 patient developed local skin infection – required surgical excision
    • 1 patient had lumbar drain torn off, requiring surgical removal

Authors’ Conclusions

• In patients with aSAH, prophylactic lumbar CSF drainage to lessen the burden of secondary cerebral infarction and unfavourable outcome is warranted

Strengths

• Pragmatic as study design relies on local standard practice allowing easy translation to routine clinical care
• Meaningful patient centred outcome
• Functional outcomes determined by blinded evaluators minimising detection bias
• Included poor grades of aSAH which other trials haven’t, therefore increasing relevance for neurocritical care population

Weaknesses

• Majority of patients randomised from 2 centers
  • Possibly lower risk of adverse outcomes if lumbar drains more routinely used at these sites
• Clinicians may alter their ‘standard’ therapy based on unblinded randomisation
• Selection bias possible due to limited screening data available (eTable 2)
  • Certain patient groups may have been excluded that would have changed outcome (e.g. 63 excluded for poor grade)
• 35 patients in lumbar drainage group were not treated as per protocol
  • 13 had no drain inserted, 22 had drain inserted but no relevant drainage performed
  • However, no difference in sensitivity analysis (per protocol and as treated)
• Potential for time-lag bias (trial enrolled from 2011 to 2016)

The Bottom Line

• Use of lumbar CSF drainage may improve neurological outcomes in patients with aSAH in addition to standard care
• The use of lumbar CSF drainage must be a multidisciplinary decision; whilst acknowledging the potential high-risk nature of the intervention, I will await further studies confirming the potential benefits of this intervention

External Links

• article Effectiveness of Lumbar Cerebrospinal Fluid Drain Among Patients With Aneurysmal Subarachnoid Hemorrhage 

Metadata

Summary author: Bridget Jones
Summary date: 23rd January 2024
Peer-review editor: George Walker

Picture by: Raman Oza / Pixabay