# Outcomes and risk factors of hemorrhage in patients with resected brain metastases

**Authors:** Melisa S. Guelen, Kiarash Ferdowssian, Niklas Jung, Hava N. Celik, Andrea Dell'Orco, Semil Eminovic, Anton Früh, Majd Samman, Güliz Acker, Arend Koch, Helena Radbruch, Michael Scheel, Mike P. Wattjes, Julia Onken, Peter Vajkoczy, Nils Hecht, Jawed Nawabi, David Wasilewski

PMC · DOI: 10.1002/ijc.70250 · International Journal of Cancer · 2025-12-07

## TL;DR

This study examines brain metastases with hemorrhage, finding that tumor size and cancer type influence bleeding risk, while survival is more affected by tumor burden and treatment.

## Contribution

Identifies tumor volume and primary tumor histology as key predictors of hemorrhage in brain metastases, challenging assumptions about antithrombotic use.

## Key findings

- ICH is associated with larger tumor volume and melanoma, but not antithrombotic use.
- Hemorrhage does not independently predict worse overall survival.
- Improved survival is linked to better functional status, solitary metastases, and adjuvant treatments.

## Abstract

Brain metastases (BrMs) may present with intralesional or intracranial hemorrhage (ICH), yet risk factors and outcomes remain unclear. This monocentric cohort study at Germany's largest neurosurgical clinic included 973 adults undergoing BrM resection (2010–2024), with histopathologically confirmed etiologies and known tumor burden. Based on pre‐operative CT or MRI, 880 patients were categorized as non‐hemorrhagic (non‐hBrM), presenting with intralesional hemorrhage (hBrM), or with ICH of ≥30 mm diameter (ICH‐BrM). Risk factors for hBrM and ICH‐BrM were assessed, and overall survival (OS) and progression‐free survival (PFS) were analyzed using Kaplan‐Meyer methods. Of 880 patients, 560 (63.6%) were non‐hBrM, 243 (27.6%) hBrM, and 77 (8.8%) ICH‐BrM. ICH‐BrM had larger tumor volume (21 cm3, IQR 13–34) than hBrM (14 cm3, IQR 6–28) and non‐hBrM (12 cm3, IQR 6–21) (p
adjust = .017), correlated with lower post‐op Karnofsky index (p
adjust = .047), dsGPA score (p
adjust = .032), and more BrMs (p
adjust = .004). Pre‐operative antithrombotic use did not differ between groups (p
adjust = .32). Melanoma was more common in hBrM (27.8%) and ICH‐BrM (38.0%), predicting ICH (OR 2.95, p < .001) along with NSCLC (OR 1.64, p < .001). ICH did not independently predict worse OS (HR 1.23, p = .38). Worse OS was linked to larger tumor volume (HR 1.35, p = .002), extracranial metastases (HR 1.77, p < .001), and older age (HR 1.53, p < .001), while KPS >80% (HR 0.77, p < .01), solitary BrM (HR 0.62, p = .002), and adjuvant treatments (p < .001) predicted improved OS. ICH is associated with larger tumors and melanoma but is not an independent OS predictor. Tumor burden, extracranial metastases, and adjuvant treatments drive BrM survival.

Brain metastases may present with hemorrhage, yet risk factors and clinical outcomes remain unclear. Based on a cohort of 880 patients with resected brain metastases, this study identified tumor volume and primary tumor histology (melanoma and non‐small cell lung cancer) as key predictors of hemorrhage. Notably, antithrombotic medication was not a significant risk factor. Patient functional status and adjuvant treatments predicted improved overall survival. Hemorrhage did not independently predict overall survival, emphasizing the dominance of tumor burden, extracranial metastases, and adjuvant treatments in driving outcomes. These findings provide critical insights for guiding the clinical management of hemorrhagic brain metastases.

## Linked entities

- **Diseases:** melanoma (MONDO:0005105), non-small cell lung cancer (MONDO:0005233)

## Full-text entities

- **Diseases:** Melanoma (MESH:D008545), BrMs (MESH:D001932), Tumor (MESH:D009369), hemorrhage (MESH:D006470), ICH (MESH:D020300), metastases (MESH:D009362)
- **Chemicals:** antithrombotic (-)
- **Species:** Homo sapiens (human, species) [taxon 9606]

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC12922642/full.md

## Figures

2 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12922642/full.md

## References

28 references — full list in the complete paper: https://tomesphere.com/paper/PMC12922642/full.md

---
Source: https://tomesphere.com/paper/PMC12922642