Divani A, Liu X, Petersen A, et al. The magnitude of blood pressure reduction predicts poor in-hospital outcome in acute intracerebral hemorrhage. Neurocrit Care (2020).
Reviewed by Kyle Hobbs, MD
Read the article.
While elevated systolic blood pressure (SBP) after ICH is associated with worse outcome, the optimal systolic blood pressure target in the acute phase is unknown.
This study was a retrospective analysis of 10 years of admission data on spontaneous ICH patients from two hospitals in Minnesota. Patients who died <24 hours after admission and had >10 missing SBP readings or who had DNR directives were excluded.
Hourly BP was recorded for the initial 6 hours, then every 2 hours from 8-24 hours post-admission (16 measurements total). Hematoma characteristics including intraventricular hemorrhage (IVH), midline shift (MLS), hydrocephalus, herniation, multisite hemorrhage and irregularity of hematoma shape were recorded. Hematoma expansion (HE) occurred when hematoma volume increased >6 mL or had a relative increase of >33% on follow-up CT.
Primary outcome was the modified Rankin Scale (mRS) score at discharge; secondary outcomes were in-hospital death (mRS 6) and HE. Statistical analysis included a functional principal component analysis (FPCA) applied to the SBP trajectories, with median and interquartile ranges computed for variables, which were assessed in a logistic regression analysis. A linear regression model converted the second SBP featured from FPCA into the magnitude of SBP reduction in the first 6 hours. Tertile cutoffs for SBP and hematoma volume were used (low, medium, high initial SBP and small, medium and large volume). Multivariate logistic regression models were used to assess associations between SBP reduction and the primary outcome.
757 patients were included in the analysis. Age, admission GCS, history of coagulopathy or cardiovascular disease/stroke, admission glucose, comfort care directive, IVH, hydrocephalus, midline shift, multisite hemorrhage, irregular hematoma shape, initial hematoma volume, HE, mean SBP and SBP reduction were all significantly associated with discharge mRS.
Mean SBP reduction over the first 6 hours was significantly greater with higher initial SBP values but was not significantly different with larger initial hematoma volumes. Mean SBP reduction was also significantly larger in patients with a history of untreated hypertension than in patients who were normotensive or had a history of treated hypertension.
Risk of poor in-hospital outcome varied according to initial hematoma volume and degree of SBP reduction, with increased probability of good outcome (mRS 0-2) with smaller SBP reduction in small (<7.42 mL) and medium-sized (7.43 to 30.46 mL) hematomas. Smaller SBP reduction decreased the probability of severe outcome (mRS 5-6) in large hematomas (> 30.47 mL). Greater SBP reduction was associated with lower risk of death for small and medium-sized hematomas, but large hematomas had a higher risk of death. Greater SBP reduction was also associated with decreased hematoma expansion regardless of initial SBP or hematoma volume. Multivariate regression models suggested that higher SBP and larger initial hematoma volume were associated with a shift toward worse outcomes. SBP reduction between 40-60 mm Hg or >60 mm Hg conferred significantly increased risk of poor outcome compared to SBP reduction <20 mm Hg, but only patient with SBP reduction >60 mm Hg had increased risk of death.
This observational study used FPCA to identify achieved mean SBP level over 24 hours and SBP reduction within the first 6 hours post-ICH as having significant impact on outcome at hospital discharge.
It is interesting that smaller SBP reduction was associated with better outcomes in small and medium-sized hematomas, as well as with a lower probability of severe outcome in large hematomas. A negative effect was seen on hospital outcome with larger reductions in SBP. Increased probability of death but decreased risk of hematoma expansion was seen with large (>60 mm Hg) reductions in BP regardless of hematoma volume or initial SBP, suggesting aggressive BP reduction may cause other detriments that overwhelm the benefits of limiting HE, or alternately, that perhaps large hematomas have already undergone HE, reducing the benefit of aggressive BP reduction.
Lower risk of death was seen in small hematomas that underwent large SBP reduction, suggesting a beneficial effect in lowering BP to reduce HE in hematomas that possibly have not yet expanded. The heterogeneity in outcomes seen depended highly on degree of SBP reduction and initial hematoma size, suggesting that a more individualized approach to SBP reduction may be more beneficial than aiming for a uniform SBP target after ICH.
Limitations of this study included its retrospective study design, as well as exclusion of subjects who died within 24 hours of admission or had comfort care directives, which may have introduced selection bias. Outcomes were limited to hospital discharge, so it is unknown if these results would be sustained over longer terms. The association between initial hematoma volume and degree of SBP reduction over the first 6 hours after ICH suggest that future protocols may benefit from a tailored approach to SBP reduction, taking into account initial ICH volume and baseline SBP, but a prospective trial is needed to further elucidate the relationships between BP reduction and long-term outcome.