# Wake-up intracerebral hemorrhage: hematoma expansion and outcomes

## TL;DR

This study finds that wake-up intracerebral hemorrhage has similar risks of bleeding expansion and outcomes as known-onset hemorrhage.

## Contribution

The study provides new evidence that wake-up ICH should not be excluded from trials targeting hematoma expansion.

## Key findings

- Wake-up ICH patients had similar rates of hematoma expansion as known-onset ICH patients.
- Wake-up ICH showed no significant differences in 3-month outcomes compared to known-onset ICH.
- Wake-up ICH patients had higher rates of dementia and anticoagulant use but similar bleeding volumes.

## Abstract

While understudied, wake-up intracerebral hemorrhage (WU-ICH) is not uncommon (8.8–20.3% of ICH patients). Since the risk of hematoma expansion (HE) decreases as time passes, an uncertain onset time in WU-ICH may influence the risk of in-hospital HE and the potential effects of HE-preventive treatments. We aimed to evaluate HE and outcomes in WU-ICH compared to known-onset ICH.

We included ICH patients admitted to the Karolinska University Hospital from 2016 to 2022, comparing WU-ICH vs. known-onset ICH regarding baseline characteristics, HE, and outcomes.

Of 763 patients, 147 (19%) had WU-ICH and 616 (81%) had known onset, median (IQR) last-known-well to hospital time 9.6 h (5.9–12.2 h) vs. 1.3 h (0.9–2.0 h). WU-ICH patients more often had dementia (15% vs. 5%, p < 0.001), oral anticoagulants (26% vs. 16%, p = 0.005), and pre-stroke modified Rankin Scale 3–5 (24% vs. 15%, p = 0.01). Baseline ICH volume was 14 mL (6–35 mL) vs. 13 mL (5–34 mL). Among patients who underwent CT angiography at admission, 15% of WU-ICH vs. 27% of known-onset ICH had spot signs (p = 0.002). Of patients with CT follow-up <72 h, HE occurred in 24/77 (31.2%) in WU-ICH, and 123/356 (34.6%) in known-onset ICH, p = 0.57. Wake-up onset was not associated with HE in multivariable analysis, adjusted OR = 0.79 (95% CI 0.43–1.42). Analysis of the 3-month modified Rankin Scale showed no differences (median 4 vs. 4), unadjusted p = 0.35 and adjusted p = 0.78.

WU-ICH had a similar risk of HE and similar 3-month outcomes as known-onset ICH. Excluding WU-ICH from future trials targeting HE may be unwarranted.