Lecture Notes: Predictors of Hematoma Expansion

Introduction

• Recent analyses presented at the International Stroke Congress.
• Focus on predictors of hematoma expansion and the impact of certain variables on treatment efficacy.

Study Details

• Participants: Patients with intraparenchymal or intraventricular hemorrhage.
• Exclusion: Patients with subdural and subarachnoid hemorrhages.
• Sample size: Approximately 460 participants.
• Baseline demographics were evenly distributed.

Primary Outcome

• Defined as hematoma growth from baseline to 12 hours.
• Absolute cutoff: 12.5 cc or greater.
• Proportional cutoff: 35% or greater.
• Predictive values:
  • 12.5 cc has 90% positive predictive value for poor outcomes at 90 days (MRS 4-6).
  • 35% cutoff has about 70% positive predictive value for MRS 4-6 at 90 days.

Logistic Regression Models

Model 1

• Variables: Treatment with Andexanet, time from symptom onset to treatment, baseline anti-faculty activity, and baseline ICH volume.

Model 2

• Additional metric: Average pre-scan hematoma growth (ultra-early hematoma growth).

Findings

• Inverse association between time from symptom onset to treatment and hematoma growth.
• Positive association with baseline ICH volume and hematoma growth.
• Strongest predictor: Baseline hematoma growth rate.
• Anti-factor activity significance diminished possibly due to collinearity with other variables.

Risk Assessment

• Linear relationship with baseline ICH volume and risk of expansion.
• High risk in the highest quartile of baseline ICH volume (22 cc's) and pre-scan growth rate (11.4 cc or greater).
• First quartile of time from symptom to treatment (<3.3 hours) showed a ~50% risk of expansion.
• Even low-risk groups had a ~20-25% expansion rate.

Effect of Treatment

• Consistent proportional reduction of hematoma expansion with Andexanet.
• Greater absolute benefit in highest risk groups:
  • 25 patients benefited per 100 treated in highest quartiles.
  • Number needed to treat: 3 to 4 to prevent expansion.
• No relationship with excess thrombotic events.

Conclusion

• Number needed to treat across all quartiles was less than 10.
• Amplified benefit in highest quartile groups for baseline ICH volume and growth rate.
• Number needed to harm remained constant around 26.
• Future analyses will focus on clinical outcomes vs. ischemic stroke consequences.