Staged stereotactic catheter drainage with urokinase thrombolysis versus neuroendoscopic evacuation for moderate-volume hypertensive basal ganglia intracerebral hemorrhage: an exploratory retrospective cohort study

The optimal minimally invasive strategy for selected patients with moderate-volume deep intracerebral hemorrhage (ICH) remains uncertain. We compared staged stereotactic catheter drainage with urokinase thrombolysis and neuroendoscopic evacuation in patients with 20–40 mL hypertensive basal ganglia ICH. We retrospectively analyzed 48 patients treated from January 2022 to December 2024: 21 underwent stereotactic catheter drainage followed by once-daily urokinase and 27 underwent neuronavigation-assisted neuroendoscopic evacuation. None of the patients had intraventricular extension, and all underwent a first postoperative CT within 24 h of surgery. In the catheter group, this scan was obtained before the first urokinase administration. The main comparative clinical outcome was the 3-month modified Rankin Scale (mRS). A proportional-odds model adjusted for age, admission Glasgow Coma Scale score, and baseline hematoma volume. Catheter-lysis had shorter operative time (44 [39–48] vs. 117 [106–133] min), less blood loss (12 [10–14] vs. 150 [100–200] mL), and lower hospitalization cost (CNY 31,005 [26,048–38,909] vs. 59,655 [50,038–73,495]; all P < 0.001). Neuroendoscopy achieved greater early postoperative hematoma reduction on the first postoperative CT: residual hematoma volume was 4.0 [3.0–10.7] versus 13.5 [12.0–21.3] mL, and hematoma clearance was 81.3% [59.5–88.7%] versus 40.3% [18.6–55.8%] (both P < 0.001). Unadjusted 3-month mRS favored catheter-lysis (4 [3–5] vs. 5 [4.5–5], P = 0.007), but the adjusted association was not statistically significant (common OR for a shift toward a worse mRS category, 0.33; 95% CI 0.09–1.21; P = 0.095). Catheter-lysis was associated with lower operative burden and cost, whereas neuroendoscopy achieved greater hematoma reduction on the first postoperative CT. Functional differences were uncertain after adjustment. These findings are exploratory and should not be interpreted as evidence of robotic-platform superiority. Not applicable; this retrospective observational cohort study was not a clinical trial.