Abstract P447: Intracranial Pressure and Cerebral Perfusion Pressure Monitoring in Spontaneous Intracranial Hemorrhage: A Secondary Analysis of the Mistie III Trial

Abstract

Introduction: Intracerebral hemorrhage (ICH) management guidelines recommend maintaining intracranial pressure (ICP) <20 and cerebral perfusion pressure (CPP) between 50-70 mmHg. We did subgroup analyses of MISTIE III trial to explore whether minimally invasive surgery (MIS) improves ICP or CPP and whether thresholds are associated with long term outcomes. Methods: MISTIE III was a randomized clinical trial including 499 patients with spontaneous ICH randomized to MIS+Alteplase or standard medical care (SMC). Primary outcomes were any threshold event of ICP >20 and CPP <60/70 mmHg. Secondary outcomes were poor modified Rankin Scale at one year and mortality at 30/365 days. We used multivariable models to investigate factors associated with ICP/CPP events and outcomes. Results: Of 72 patients with ICP monitored for median 92 (72-96) hours, 31 (43.1%) had at least one ICP reading >20 and 52/35 (72.2/34.7%) had at least one CPP reading <70/60 mmHg. Lower intraventricular hemorrhage volume and SMC group were associated with having any ICP threshold event >20 and CPP event <70 mmHg whereas CPP<60 mmHg was associated with end of treatment (EOT) ICH volume, hydrocephalus on diagnostic CT and no prior antiplatelet agent use. On adjusted analyses, percentage of ICP readings >20 were significantly less likely in patients undergoing MIS vs SMC (Coefficient -0.79, 95% Confidence Interval [CI] [(-)1.46-(-)0.11]; p=0.02). Percentage of CPP readings <70 were significantly less frequent in MIS group (Coefficient -1.59 [(-)2.58-(-)0.59]; p=0.002). Patients who underwent successful MIS with EOT ICH volume <15mL also had significantly lower percentage of readings at ICP<20 (p=0.02), and CPP<70 (p=0.05). Lower percentage of CPP readings <60 mmHg was independently associated with lower mortality at 30 and 365 days (p=0.02 and 0.04) and CPP <70 was associated with lower one-year mortality (p=0.04). There were no significant associations with one-year functional outcome. Conclusion: Elevated ICP and inadequate CPP are not infrequent during ICP monitoring for large ICH. Burden of low CPP events predict higher short and long term mortality, but not functional outcomes. CPP may be more significant than ICP. MIS appears to mitigate ICP and CPP threshold events.