Dynamics of intraventricular hemorrhage in patients with stroke due to spontaneous intracerebral hemorrhage

Abstract

Intracerebral hemorrhage (ICH) is a devastating type of stroke that represents a significant global burden of disease. It has a 50% mortality rate at 30 days and leaves 75-80% of patients with some form of long term disability. Extension of the parenchymal hematoma into the ventricles, or an intraventricular hemorrhage (IVH), is a secondary event that carries significant co-morbidities in at least one third of all ICH. IVH is an independent and robust risk factor for poor clinical outcomes that increases mortality and risk of permanent functional impairment. The presence of IVH has a stronger negative prognostic index than other known predictors of poor outcome in ICH, including larger baseline ICH volume, worse Glasgow Coma Scale and increasing age. IVH thus appears associated with the subgroup of patients with the greatest severity of ICH. In addition to epidemiological data of the prognostic significance of IVH, animal models have elucidated a biological basis for the increased occurrence of early neurological deterioration, coma and other complications. A combination of the mechanical effect of ventricular clot increasing resistance to flow of cerebrospinal fluid and inflammation from the toxicity of blood degradation products causes the hydrocephalus and dangerous increases in intracerebral pressure that are commonly seen with the presence of IVH. Further complications can arise from these events including mass effect on surrounding brain parenchyma and herniation. IVH produces more extensive damage than the local effects of ICH, and in some instances may potentiate the injury caused by ICH. This pathophysiological understanding of the effect of IVH has provided a rationale for novel techniques of early and safe removal of ventricular clot. The Clot Lysis Evaluating Accelerated Resolution of Intraventricular Hemorrhage (CLEAR) trials are investigating the efficacy of catheter-guided intraventricular clot lysis and drainage using recombinant tissue plasminogen activator (rtPa) with preliminary results already showing an acceptable clinical safety profile. The promise of randomised evidence from the final phase CLEAR III trial demonstrating the risks and benefits of rapid clot clearance has prompted interest in determining a clinically significant IVH volume and its risk of poor outcome that would warrant such an invasive intervention. On the other hand, this raises questions about the efficacy of less invasive forms of IVH management that target risk factors of IVH presence and growth. Before any meaningful and marked change in clinical practice can occur, there needs to be more robust epidemiological data of IVH, clinical outcomes and associated risk factors from a large scale randomised trial. This thesis therefore aimed to firstly examine the risk associations of IVH and clinical outcomes, and characterise the precise relationship between IVH volume and poor outcomes; secondly, to determine whether early intensive blood pressure (BP) lowering has a significant impact on IVH growth; and thirdly, to analyze how these findings are beneficial to current and future management of IVH in acute spontaneous ICH. The findings in this thesis are based upon the large clinical datasets of the Intensive Blood Pressure Reduction in Acute Cerebral Hemorrhage Trials (INTERACT 1 and 2). Following an initial pilot phase in INTERACT1 of 404 patients, INTERACT2 evaluated the effectiveness of early intensive BP lowering treatment in 2839 patients with ICH who were recruited from over 120 centres in 20 countries during 2008-2012. Eligible patients (age ≥18 years) with spontaneous ICH confirmed by computerised tomography (CT) within 6 hours of onset, were randomized to either a regime of early intensive BP lowering (140mmHg systolic target) or conservative BP lowering (180mmHg systolic target) after providing informed consent according to local ethics committee approvals. A subset (n~1000) of all randomized patients (N=3200) in both INTERACT1 and 2 had both baseline and repeat 24 hour CT scans, allowing an assessment of the dynamics of IVH. Chapter 2 elucidated the pattern of association between IVH volume and clinical outcomes among patients of the INTERACT2 study. Linear and logistic regression models were used to analyze the risk of death and major disability at 90 days in 740/2613 (28%) patients with baseline IVH. The odds risk of poor outcomes, defined as death or major disability, in patients with IVH was two times that of ICH-only patients (95% confidence interval 1.67-2.38, p<0.01). 67% of patients with IVH had poor outcome whereas only 49% of ICH- only patients had the same. There was a strong and linear relationship between IVH volume and poor outcomes in ICH, and tertiles analysis of IVH volume suggested 5 to 10mL as a significant and clinically meaningful threshold for increased mortality and poor outcomes, though not for major disability alone. Chapter 3 looked at the impact of early and intensive BP lowering on IVH growth in the first 24 hours of patients in a pooled CT sub-studies analysis of INTERACT1 and 2. The association of guideline vs intensive BP management and categories of achieved systolic BP over 24 hours with IVH growth was assessed by analysis of covariance (ANCOVA). No significant reduction in IVH growth was evident in 228 patients who underwent intensive BP lowering compared to guideline-recommended management (1.6mL vs 2.2mL respectively, p=0.56). There was no clear benefit of intensive BP reduction for IVH growth in patients with ICH and secondary IVH despite a trend of greater IVH growth with higher achieved systolic BP over 24 hours (p=0.26). Finally, Chapter 4 discussed the implications for current and future IVH management of all the results from previous Chapters. Although there is still a lack of robust and randomized evidence for the specific management of IVH, our findings reaffirm the importance of recognising IVH early for predicting and potentially altering poor outcomes. Despite considerable emphasis on baseline IVH in the literature, we urge for more focus on managing IVH growth and delayed IVH (dIVH), both of which were stronger predictors of worse prognosis. Having established the risk associations of IVH volume and outcome, as well as that for dIVH at 24 hours, these factors should be included in existing outcomes scoring systems like the ICH score and IVH score. Decisions on the need for and timing of clinical management should be guided by the risk of poor outcomes, which were characterized in Chapter 1. In particular, the clinically significant threshold of 5-10mL of IVH would be pertinent to large scale and effective undertaking of intraventricular clot lysis upon the completion of CLEAR III. Increased understanding of IVH, its risk factors and the implication for clinical outcomes allows clinicians and researchers to better treat the severe subpopulation of patients with ICH and concurrent IVH.