Association of prophylactic mannitol administration with perihemorrhagic edema and long-term outcome in patients with intracerebral hemorrhage

Perihemorrhagic edema (PHE) is an independent predictor of functional outcome after intracerebral hemorrhage (ICH). Although smaller studies have shown an association between prolonged prophylactic mannitol administration (PPMA) and PHE volume reduction, its therapeutic effect remains unclear. We included patients with spontaneous ICH from the UKER-ICH database. The primary outcome was defined as the change in PHE volume from admission to its maximum extent (PHE growth). We used linear regression models adjusted for peak ICH volume to estimate associations of PPMA with PHE, and multivariable logistic regression models to evaluate functional outcome (day-90; modified Rankin-Scale 4–6 = poor). 826 patients were included. Ninety-five patients received mannitol. The treated patients had larger median peak ICH volumes (42 (IQR 19.2–70.4) vs. 11.7 (IQR 4.6–27.7) mL, p < 0.001). Linear regression demonstrated an association between PPMA and increased PHE growth and peak PHE volume (adjusted b = 6.2, 95%CI 2.16–10.63 and b = 8.4, 95%CI 3.16–13.6, respectively). After PS-matching (n = 116), the peak ICH volume was evenly distributed, while PPMA was associated with larger median peak PHE volumes (45.7 (IQR 24-78.2) vs. 35.2 (IQR 19.7–52.7) mL, p < 0.05) and growth (21.5 (IQR 9.2-42-8) vs. 11.9 (IQR 1.4–25.8) mL, p < 0.05). PPMA was associated with poor outcome (adjusted OR = 2.09, 95%CI 1.01-4.00; PS-matched cohort (n = 114): 45 (79%) vs. 36 (65%), p = 0.097 (trend)). We found no association between PPMA and reduced PHE volume or improved outcomes, and cannot exclude potential harm. However, we also cannot exclude a potential benefit of short-term mannitol administration up to 72 h after onset in reducing PHE volume. https://www.clinicaltrials.gov/study/NCT03183167 (2017-06-09)