• Ischemia Clinical Trials

Patients with symptomatic intracranial hemorrhage (sICH) have an increased risk of a poor or fatal outcome [Strbian D et al. Neurology 2011]. However, the direct comparison of sICH rates between different thrombolysis studies is complicated by varying definitions of sICH [Gumbinger C. et al. Stroke 2012].

Definitions include those from The National Institute of Neurological Disorders and Stroke-tPA study (NINDS) [The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. N Engl J Med 1995], the second European-Australasian Acute Stroke Study (ECASS-II) [Larrue V et al. Stroke 2001], ECASS-III [Hacke W. et al. N Engl J Med 2008], and The Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST) [Wahlgren N. et al. Lancet 2007]. Outcomes differ according to clinical, radiological, and relational criteria.

Neal M. Rao, MD, University of Colorado School of Medicine, Denver, Colorado, USA, presented results from a study on the most accurate definition of clinically relevant hemorrhagic transformation (HT) after thrombolytic therapy for stroke with IV tissue plasminogen activator (tPA).

The specific aim of this study was to determine which definition of sICH best identifies hemorrhages that alter final patient outcomes after administration of intravenous tPA in acute stroke. Analysis was based on the NINDS database, which defines sICH as any hemorrhagic transformation that is temporally related to any worsening.

Methods included an analysis of candidate definitions—ie, radiological (any radiological hemorrhage or parenchymal hematoma [PH])—and clinical-radiological criteria (NINDS-tPA Study, ECASS-II, and modified SITS-MOST: PH and ≥4 National Institutes of Health Stroke Scale [NIHSS] worsening).

Clinically relevant hemorrhages were defined as those that altered final outcome. A predictive model from the placebo group was derived, and outcomes with tPA were compared with predicted outcomes without tPA using a modified Rankin Scale (mRS).

The data of 312 patients who were treated with IV tPA were analyzed; 48 patients (15.4%) experienced any radiological intracranial hemorrhage (ICH). Hemorrhage frequency varied by definition (6.4%, any [PH]; 6.4%, NINDS-tPA; 5.1%, ECASS-II; and 1.9%, modified SITS-MOST). ECASS-II sICH patients had worse actual (with tPA) versus predicted (without tPA) outcomes. The mean final mRS was 5.6 (observed) versus 3.5 (predicted); death occurred in 75% (observed) versus 25.4% (predicted) of patients.

Radiological hemorrhage patients who did not meet ECASS sICH criteria showed no difference between actual and predicted outcomes. Mean final mRS was 4.2 (observed) versus 4.0 (predicted); death occurred in 35% (observed) versus 35.1% (predicted) of patients. Table 1 shows actual and predicted mean mRS and mortality with and without the five definitions (any radiographic, PH, NINDS, ECASS-II, and SITS-MOST).

Study limitations were: only a subset of the proposed definitions of sICH was analyzed, data from the NINDS trials may not fully reflect contemporary practice, and there were a small number of patients in the NINDS dataset.

The authors concluded that the ICH classification that best identifies clinically relevant hemorrhages that alter final global disability and fatal outcome is any radiological HT that is associated with ≥4 early NIHSS worsening. They also determined that asymptomatic hemorrhages under this definition have no adverse impact on final outcomes.

• © 2012 MD Conference Express®