ENRICH: Early Minimally Invasive Removal of ICH

Trial of Early Minimally Invasive Removal of Intracerebral Hemorrhage

Pradilla. NEJM 2024; DOI: 10.1056/NEJMoa2308440

Clinical Question

  • In patients with spontaneous intracranial haemorrhage (ICH), does the use of early minimally invasive surgery + standard management compared to standard management alone, result in better functional outcomes at 180 days?

Background

  • Spontaneous ICH is a condition with high mortality, poor functional outcomes in survivors and high costs of care
  • The mainstay of standard medical management for ICH has been reversal of coagulopathy, blood pressure control and best supportive care
  • While showing mortality benefit, surgical intervention with craniectomy (STITCH, STITCH II) and minimally invasive techniques (MISTIE III) has failed to show long term benefits in functional outcomes
  • Data from these trials showed a trend towards benefit with early surgical intervention in superficial lesions and with greater clot evacuation
  • The ENRICH trial was designed to investigate any benefit of early minimally invasive surgery trans-sulcal parafascicular surgery (MIPS) in superficial ICH

Design

  • Multi-centre, two-arm, randomised, adaptive design
  • 1:1 block randomization stratified by:
    • ICH location (anterior basal ganglia vs lobar)
    • Index GCS (<9 or ≥ 9)
  • Subjects followed to 6 months or until death
  • Utility weighted mRS used:
    • Weights of 1.0, 0.91, 0.76, 0.65, 0.33, 0.0, and 0.0 were applied to the 7 levels from mRS 0 (best) to 6 (worst) respectively
      • This means there are unequal gaps between different levels e.g. severe disability and death (mRS 5 and 6) are scored the same, but there is a 0.15 difference between mRS 2 and 3
  • Planned adaptive sample size 150 – 300 patients
  • Adaptive design allowing longitudinal modelling for interim decision to predict 180-day outcome using 90-day mRS
    • Interim analysis at 150 patients and then in blocks of 25 to determine if early stopping rules are met:
      • If accrual stopped due to futility in one site (anterior basal ganglia or lobar) then this would lead to enrichment for the other site
    • Pre-specified stopping rules for both futility and success
    • Primary outcome defined as between group mean differences in utility weighted mRS (H1: Δ> 0)
      • To demonstrate superiority, a Bayesian model-based posterior probability of MIPS benefit (i.e., Δ> 0) was set to be ≥0.975
    • Non-informative prior used for Bayesian analyses
    • Formal statistical consulting service (Berry Consultants) used
  • Scores on mRS obtained by site personnel with structured interviews with patients and caregivers
    • In order to mask primary endpoint audio-recordings were uploaded with redaction of any information about group assignment and reviewed by a blinded neuropsychologist
  • Written consent obtained prior to enrolment

Setting

  • December 2016 – August 2022
  • 37 hospitals in the USA
    • 59 treating Neurosurgeons

Population

  • Inclusion:
    • Aged 18 – 80
    • Acute spontaneous, primary ICH with volume of 30-80ml
    • Intervention within 24hrs of last known well
    • GCS 5-15
    • Previously mRS 0 -1 before ICH occurred
  • Exclusion:
    • Recurrent ICH or other underlying vascular abnormality
    • Primary thalamic or infratentorial intraparenchymal haemorrhage
    • NIHSS <5
    • IVH >50% of lateral ventricles
    • Uncorrected coagulopathy or a known clotting disorder
    • Anticoagulants not rapidly revisable or need for anticoagulation within 5 days of index admission
    • Not expected to recover, DNR, life expectancy < 6 months
  • Comparing baseline characteristics of surgery vs. control group
    • Age: 64 vs 62 years
    • Female Sex: 48 vs 52%
    • Median score on NIH stroke scale: 16 vs 18
    • GCS at randomisation:
      • 4 – 8: 17% vs 19%
      • 9 -14: 83% vs 81%
    • Median haematoma volume: 54 vs 55 ml
    • Intraventricular haemorrhage present: 43% vs 39%
    • ICH location:
      • Anterior Basal Ganglia (ABG): 32 % vs 29%
      • Lobar 68% vs 71%
    • Median ICH score: 2 vs 2
    • Median APACHE II score: 12 vs 12
    • Medical History:
      • CVD: 82% vs 78%
      • CNS disease: 30 % vs 28%
      • Endocrine or renal disorder: 28% vs 26%
      • Smoker (current/former): 45% vs 40%
    • Time from last known well to randomisation:
      • ≤8hrs: 28% vs 25%
      • ≤12hrs: 49% vs 46%
  • 300 patients enrolled
    • 11603 assessed for eligibility
      • 10611 did not meet inclusion criteria / met exclusion criteria
      • 692 excluded with ABG hemorrhage after adaptation of enrollment criteria

Intervention

  • Early ICH evacuation (within 24hrs, ideally <8hrs) using minimally invasive trans-sulcal para fascicular surgery (MIPS)+ standard management
    • Standardized surgical approach with training provided
    • Imaging guided
    • BrainPath and Myriad device for minimally invasive technique (manufactured by NICO)
    • Video recording and independent review of surgeries at each site
  • Pre-specified futility criteria for ABG haemorrhage met at the interim analysis for 175 patients
    • Following this only lobar haemorrhages enrolled
  • Median time from last known well to surgery was 16.8 hours

Control

  • Standard management alone

Management common to both groups

  • Based on AHA/ASA guidelines
    • Aggressive resuscitation
    • Blood pressure control
    • Correction of bleeding diathesis
    • Traditional surgical techniques permitted (EVD, craniectomy etc.)

Outcome

  • Primary outcome:
    • Mean score on utility weighted modified Rankin scale at 180 days:
      • 0.458 (surgical group) vs 0.374 (control group)
      • Difference of 0.084 (95% Cr I 0.005 to 0.163) and posterior probability of superiority 0.981
    • This benefit was attributable to ICH in lobar location (Figure S2):
      • Anterior Basal Ganglia: 0.340 (surgery) vs 0.381 (control)
        • Difference of -0.013 (95% Cr I -0.147 to 0.116)
      • Lobar: 0.513 (surgery) vs 0.371 (control)
        • Difference of 0.127 (95% Cr I 0.035 to 0.219)
    • The benefit for intervention on mRS was seen at discharge / 7 days, 30 days, 90 days and 180 days
  • Secondary and Safety Outcomes:
  • Comparing surgery vs. control group
    • Significantly greater in surgery group:
      • Change in haematoma volume: -43.9 vs 4.0 ml
        • Difference -47.9 (-53.6 to -42.4 mls)
    • No significant difference in
      • Score of ≤3 on mRS at 180 days: 50.3 vs 41.0%
        • Difference 9.2% (95% Cr I -2.0 to 20.3)
    • Significantly less in surgery group
      • ICU Length of Stay: 6.9 vs 9.7 days
        • Difference -2.8 days (95% Cr I -4.5 to -1.1 days)
      • Hospital Length of Stay: 14.9 vs 18.1 days
        • Difference of -3.1 days (95% Cr I -5.9 to -0.4 days)
      • One or more serious adverse events: 63.3 vs 78.7%
        • Difference -15.3% (95% Cr I -25.4 to -5.2%)
  • Exploratory outcomes:
    • Decreased duration of mechanical ventilation in surgery group: 5.3 vs 8.8 days
      • Difference -3.5 days (95% Cr I -5.7 to -1.5 days)
    • Decreased rates of decompressive hemicraniectomy in surgery group: 3.3 vs 20%
      • Difference -16.6% (95% Cr I -23.9 to -9.9%)

Authors’ Conclusions

  • Among patients in whom surgery could be performed within 24 hours after an acute ICH, minimally invasive hematoma evacuation resulted in better functional outcomes at 180 days than those with guideline-based medical management. The effect of surgery appeared to be attributable to intervention for lobar haemorrhages.

Strengths

  • The rationale for the study was sound and guided by evidence and trends displayed in previous trials
  • Primary endpoint of functional outcome is a patient centred and meaningful outcome
  • Balanced baseline characteristics for the intervention and control
  • Multi-centre and randomised
  • Minimal loss to follow up (8/300)
  • Efforts to minimise bias associated with subjective reporting of mRS by including caregivers in assessment and having blinded review of the assessment by a neuropsychologist
  • Success/safety of surgical intervention in terms of reduction in haematoma volume and low rate of rebleeding
  • Pre published statistical analysis plan

Weaknesses

  • Some limitations to external validity:
    • Single country
    • Strict exclusion criteria – very specific population in terms of ICH size and location
  • Utility weighted mRS may be more patient centric given public perceptions of neurological outcomes have been shown to be non-linear with unequal gaps between health states. However:
  • Data collection around adverse events (Table S12) is very broad and range from subjective outcomes like “flu-like symptoms” and “palpitations” through to asystole
    • Whilst data collection around adverse events is important in novel interventions, the sheer number of events collected may confound the interpretation of the reported adverse event rate
  • High rates of decompressive craniectomy in the control group – this is a procedure that has been shown to increase survival but with high rates of poor functional outcomes
  • Sponsorship by NICO, the manufacture of the devices used in the trial, although no incentives provided, no involvement in data collection, analysis or trial design

The Bottom Line

  • While it represents a fractionally small population of ICH, the benefit associated with the intervention for lobar haemorrhage is functionally important
  • Decisions around surgical intervention should be based on multi-disciplinary discussions with neurosurgical specialists

External Links

Metadata

Summary author: Mitch Pincus
Summary date: 25th May 2024
Peer-review editor: George Walker

Picture by: Katrin Bolovtsova / Pexels

 

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.