부산대학교 도서관

High blood pressure (BP) is the strongest modifiable risk factor for spontaneous intracerebral haemorrhage (ICH). International guidelines about the management of high BP for primary prevention of cardiovascular disease and secondary prevention after stroke refer to mean or 'usual' BP targets for office BP monitoring (long-term visit-to-visit BP). Yet, there is mounting evidence that higher systolic BP variability (SBPV) is associated with incident cardiovascular disease and poor outcomes after ICH. In this thesis, I aimed to address three areas of uncertainty regarding the associations between SBPV and the characteristics and outcomes of patients with ICH: Question 1: Is long-term visit-to-visit SBPV before ICH associated with the characteristics of patients with incident, first-ever ICH? In the days leading up to ICH, one population-based study showed that SBP was significantly higher than average pre-morbid SBP, suggesting higher SBP levels/ higher SBPV may trigger the event. However, these findings have not been replicated in another methodologically rigorous population-based study. My systematic review did not identify any studies that assessed associations between long-term visit-to-visit SBPV before ICH onset and the characteristics of patients with ICH at diagnosis. I analysed data from the Lothian Audit of the Treatment of Cerebral Haemorrhage (LATCH), a prospective population-based study that identified adults with first-ever ICH between 1st June 2010 and 31st May 2013, inclusive, and the nested Lothian IntraCerebral Haemorrhage, Pathology, Imaging and Neurological Outcome (LINCHPIN) sub-study. I did not detect a relationship between higher SBP levels or higher SBPV during ten years before ICH and increasing temporal proximity to the time of ICH onset. Compared to patients with lower long-term visit-to-visit SBPV during ten years before ICH, patients with higher long-term visit-to-visit SBPV were more likely to have a premorbid diagnosis of hypertension, higher BP at ICH diagnosis, and evidence of greater cerebral small vessel disease (CVSD) burden on diagnostic CT brain scan. They were less likely to have CT brain characteristics that predicted the presence of cerebral amyloid angiopathy (CAA)-associated ICH in the cohort. However, after accounting for confounding factors, there were no clear and consistent associations between the degree of long-term visit-to-visit SBPV during ten years before ICH and greater CVSD burden on diagnostic CT brain scan, ICH location or the severity of underlying CAA on research autopsy. My analyses were limited by poor BP data availability, and future research may benefit from harnessing the power of electronic data linkage with primary care records. Question 2: Are short- and mid-term SBPV after acute ICH associated with clinical outcomes, independently of important confounding factors? My systematic review and meta-analysis showed that short- and mid-term SBPV after acute ICH were associated with poor medium-term (90-day) functional outcome. However, only 3 of 6 included studies accounted for confounding from mean SBP, and none accounted for the characteristics of BP-lowering interventions after acute ICH. Furthermore, evidence about the associations between short- and mid-term SBPV and other important outcomes after acute ICH was limited. I led analyses of pooled patient-level data from 3829 participants from two randomised controlled trials (RCTs) of early, intensive BP lowering after acute ICH (INTERACT2 and ATACH-II). I showed that higher short-term SBPV during 1-24 hours of BP-lowering treatment after acute ICH (<6 h of onset) was associated with poorer medium-term functional outcome, greater odds of haematoma growth and early neurological deterioration within 24 hours, and greater odds of death and serious adverse events within 90 days. These associations existed independently of other summary measures of SBP control and important confounders. In a broader meta-analysis of 6221 patients from 16 RCTs participating in the Blood pressure in Acute Stroke Collaboration (BASC), I showed that higher short- (1-24 hours) and mid-term (2-7 days) SBPV after acute ICH were associated with poorer medium-term (90-180 days) functional outcome. However, these associations varied according to the manner in which BP was lowered. The associations between higher short-term SBPV during 1-24 h after acute ICH and poorer medium-term function were strongest amongst patients who received titrated target-based BP lowering with intravenous α- and β-adrenoreceptor blockers and calcium channel blockers, and amongst patients who had their BP measured more frequently, and had larger reductions in their SBP and lower average SBP. The associations between higher mid-term SBPV during 2-7 days after acute ICH and poorer medium-term function were strongest amongst patients who were older, had lower baseline SBP, and were randomised later after symptom onset. These findings suggest that careful and sustained reductions in SBP may be beneficial after acute ICH. Future research that aims to test the effects of minimising short- and mid-term SBPV after acute ICH is required to make definite conclusions but is likely to require a very large sample size and be prone to various methodological issues. Question 3: Is it feasible to assess the associations between long-term visit-to-visit SBPV and long-term outcomes in survivors of ICH? My systematic review identified only one study that reported an association between higher long-term visit-to-visit SBPV after ICH and greater risk of recurrent ICH. The associations between long-term visit-to-visit SBPV after ICH and other serious symptomatic vascular events are unknown. My assessment of data available about long-term visit-to-visit SBPV during 7-365 days after ICH and long-term (>1 year after ICH) outcome in three prospective, population-based cohorts of ICH indicated that pre-planned analyses were unfeasible. I made this decision due to a combination of poor 1-year survival and poor BP data availability. Amongst patients with incident, first-ever ICH in Lothian, Scotland during 2019, half were eligible for long-term BP lowering after ICH. Yet, as few as 1 in 10 patients were eligible for activities that involved intensification of their long-term BP-lowering treatment and frequent BP monitoring. The majority of patients were excluded due to death and poor prognosis. Improvements in the longterm management of BP in NHS Lothian between 2010-12 and 2019 also contributed; several patients had systolic BP levels <130 mm Hg, which excluded them from the studies. My findings highlight complexities to generating evidence about interventions for secondary prevention of stroke and other serious symptomatic vascular events after ICH; future research may benefit from prospective, standardised BP data collection and broader recruitment strategies. In summary, there were no clear and consistent associations between long-term visitto-visit SBPV before ICH and the characteristics of patients with ICH at diagnosis. After acute ICH, higher short- and mid-term SBPV were associated with poor outcomes. Patients who receive titrated target-based BP-lowering interventions may benefit from careful, sustained reductions in their BP during the first week of treatment. Analyses of the associations between long-term visit-to-visit SBPV after ICH and long-term outcomes were unfeasible. Generating definitive evidence about the effects of reducing SBPV to improve outcomes after ICH is likely to require a very large sample size and be hampered by various methodological issues.