Mechanical thrombectomy (MT) has become the standard treatment for patients with acute ischemic stroke and large vessel occlusion (LVO), demonstrating superior outcomes even within extended time windows [1]. However, clinicians frequently encounter patients with concurrent intracranial hemorrhage and face therapeutic dilemmas in the absence of clear evidence-based guidelines. In this context, we distinguish intracerebral hemorrhage (ICH) from hemorrhagic transformation (HT). ICH refers to bleeding in the brain parenchyma in the absence of trauma or surgery. HT occurs within the ischemic territory and is classified by the European Cooperative Acute Stroke Study criteria as hemorrhagic infarction type 1 or 2 (HI1/HI2), or parenchymal hematoma type 1 or 2 (PH1/PH2) [2].

Patients with LVO may present with concurrent intracranial hemorrhage on initial imaging, during interhospital transfer, or after intravenous thrombolysis (IVT), and each represents a critical decision point in the treatment process. The American Heart Association and American Stroke Association (AHA/ASA) guidelines note that HI1/HI2 are typically petechial, lack mass effect, and present a low risk of deterioration, whereas PH1/PH2 are associated with worse outcomes [2]. Current guidelines provide limited recommendations for these scenarios. Consequently, clinicians may refrain from performing MT because of concerns about bleeding risk, even in the absence of supporting evidence. This is important because untreated LVO results in poor clinical outcomes. Natural history studies indicate that only 19% of patients achieve functional independence without intervention [3]. Conversely, successful MT can improve clinical outcomes by reducing disability and increasing functional independence compared with standard medical treatment [1].

This review examines the evolving evidence for individualized MT decision-making in patients with concurrent intracranial hemorrhage. Should MT be offered to patients with LVO and concurrent intracranial hemorrhage?

Concurrent intracranial hemorrhage with LVO presents in two distinct scenarios, each with unique pathophysiological mechanisms and management considerations. The first scenario involves patients who develop HT during interfacility transfer. HT occurs in 5.5% of transferred cases [4], predominantly as mild HI1/HI2. Symptomatic HT is observed in only 0.1%-0.2% of cases. Current AHA/ASA recognize that HI1/HI2 typically lack mass effect and carry a low risk of deterioration, whereas PH1/PH2 are associated with worse outcomes. The second scenario involves patients with concurrent LVO and remote ICH unrelated to the ischemic territory. This exceptionally rare condition occurs in only 0.5% of MT procedures [5] and often involves chronic microbleeds, cerebral amyloid angiopathy, or hypertensive hemorrhage. These scenarios require different risk-benefit assessments. Remote ICH requires a comprehensive assessment of hemorrhage stability and the anatomical relation with the target vessels. Treatment decisions should consider hemorrhage characteristics, patient factors, and procedural expertise rather than uniform exclusion criteria.

MT generally does not exacerbate incidental intracranial hemorrhage, including HT and ICH, although outcomes in patients with concurrent bleeding remain complex. A recent systematic review and meta-analysis of six studies by Elfil et al. [6] demonstrated that MT achieved high revascularization rates (85.3%) in patients with acute ischemic stroke and concurrent ICH. However, functional independence was achieved in only 20% of patients. This difference may reflect a complex clinical scenario in which concurrent ICH is one of several factors affecting overall prognosis. The safety of MT is based on mechanisms that differ from those of IVT. IVT alters coagulation systems and may destabilize existing hematomas, whereas MT is a localized physical intervention that removes thrombi without pharmacologic effects on coagulation. This localized nature minimizes the impact on existing hemorrhages by avoiding systemic hemostatic effects or fibrinolytic pathway activation. However, Lee et al.’s [7] large Korean registry (1,495 patients) demonstrated that previous ICH did not significantly increase the risk of symptomatic HT (odds ratio 1.08, 95% confidence interval 0.39-2.96) after IV tissue plasminogen activator. The authors concluded that previous ICH may not increase symptomatic HT risk or alter major clinical outcomes. Furthermore, they reported that IVT+MT did not increase symptomatic HT in patients with prior ICH. Multiple observational studies have demonstrated that MT does not exacerbate ICH. Tonetti et al. [8] reported eight ICH cases during MT; only one patient demonstrated radiographic worsening, and this occurred without clinical decline. This patient, who required stent placement and received an IV P2Y12 inhibitor, demonstrated slight HT at 24 hours but no mass effect or neurological deterioration. Weller et al.’s [5] German multicenter registry identified 32 patients (0.5%) with acute ischemic stroke and concurrent ICH among 6,635 MT procedures. New intracranial hemorrhages, including HT and ICH, occurred within 24 hours in 6.9% of patients with concurrent ICH and 10.2% of controls (P=0.74). Although this difference was not statistically significant, the findings suggest that MT may be safely performed in selected patients with incidental ICH. Among the 32 patients, only one case involved anatomical overlap between the ICH and vascular occlusion territory. Given the theoretically highest risk of hemorrhage expansion in same-territory ICH and LVO, generalization based on this single case is not possible. In addition, we identified three additional case reports [9-11]. Two patients presented with concurrent ICH, and one patient presented with concurrent subarachnoid hemorrhage. All patients underwent MT without worsening of the hemorrhage, and all achieved favorable outcomes. However, certain risk factors must be carefully considered. Styczen et al. [12] reported six IVT-associated hemorrhagic cases before MT. Five of these patients underwent MT, and three had PH2 HT before MT. After MT, five (83%) demonstrated aggravated HT with edema. The authors recommend special consideration in patients with preexisting HT with mass effect, such as PH2. Patients with extensive infarction also require careful risk-benefit evaluation, as those with low Alberta Stroke Program Early Computed Tomography Score (≤7) and poor collateral status have a higher risk of HT [13-15]. A 73-year-old woman presented with a right middle cerebral artery occlusion and concurrent HT. The patient was transferred for MT, and post-procedure computed tomography showed no significant expansion of the HT (Supplementary Figure 1). We assert that particular attention is required before performing MT in patients with mass effect—such as PH2—or in patients who have received IVT, have a large infarct core, or have poor collateral circulation.

In conclusion, MT does not necessarily worsen concurrent intracranial hemorrhage. Current evidence supports considering MT for patients with LVO and concurrent intracranial hemorrhage when they meet the appropriate clinical criteria.