Blood-brain barrier dysfunction and risk of dementia following blood-stream infections

Lead Research Organisation: Cardiff University
Department Name: School of Medicine

Abstract

Blood-stream infections (also known as 'septicaemia' or 'blood-poisoning') are common life-threatening infections affecting over 4,000 people in Wales (and over 50,000 in the UK) every year. They often cause sepsis which is a life-threatening condition with a 30-day mortality in excess of 15%. In those who survive the morbidity is significant with short term cognitive impairment, such as problems with memory or attention (known as delirium), frequently observed by doctors, patients and their families. Subsequent problems with memory and thinking are a frequent complaint of survivors of sepsis who often suffer from 'mental fogginess'. However, the longer term consequences and the risk of dementia are poorly understood. Greater understanding of the long-term consequences following blood-stream infections and sepsis have been identified as a research priority and as an unmet need by patient groups [UK Sepsis Trust - personal communication]. As such, there is a urgent need to determine the relationship between blood-stream infections and subsequent risk of dementia. Additionally, determining the mechanisms of brain injury associated with blood-stream infections will allow further research and interventions to prevent these complications in the future.

This project aims to determine the relationship between common blood-stream infections and subsequent dementia. This will be accomplished by using routinely collected GP and hospital data throughout Wales and linking this to blood-stream infection data collected by Public Health Wales. The risk of dementia, timing of onset and potential risk factors at a population level will be established. These data will allow an accurate estimation of the national burden of disease, guide further research and potentially provide innovative avenues for interventions to reduce dementia risk following blood-stream infection and other significant infections.

The other aim of this project is to conduct a pilot study using detailed brain scans to investigate how blood-stream infections affect the brain. Patients with blood-stream infections and healthy controls will be offered the chance to participate in a research study where they have a magnetic resonance imaging (MRI) scan near to the time of blood-stream infection and 12-18 months later. This will allow the reduction in brain size that occurs over time (known as atrophy) to be measured accurately and compared between patients and controls. Additionally, disruption to the brain's usual protective mechanisms from infection (known as the blood-brain barrier) will be measured. This will provide crucial information for future researchers to inform other brain imaging and treatment studies.

Sepsis is an increasingly common condition with a substantial mortality and long-term morbidity. This study has the potential to provide novel mechanistic insights into why sepsis frequently has such devastating long-term cognitive consequences and stimulate novel therapeutic strategies that begin to address them.

Technical Summary

Blood-stream infections (BSI) are common life-threatening infections, a leading cause of sepsis and have a 30-day mortality in excess of 15%. Sepsis-associated encephalopathy is characterised by acute neurological dysfunction in response to extra-cranial, systemic infection - occurring in up to 70% of patients with sepsis. Its mechanisms are poorly understood but there are data to suggest an increased risk of subsequent dementia.

Research questions to be answered:
1. What is the relationship between common BSI (Escherichia coli and Staphylococcus aureus) and incident dementia?
2. What are the risk factors for development of dementia following BSI with Escherichia coli and Staphylococcus aureus and do they differ by pathogen?
3. Are Escherichia coli and Staphylococcus aureus BSI associated with increased blood-brain barrier permeability?
4. Are Escherichia coli and Staphylococcus aureus BSI associated with accelerated grey and white matter atrophy?

Methods:
1. Public Health Wales Escherichia coli and Staphylococcus aureus BSI data will be linked to routinely collected GP/Hospital patient data using the Secure Anonymised Information Linkage (SAIL) databank. This will take advantage of the newly created dementia e-cohort within SAIL with ~1.2 million people and 130,000 cases of dementia. Survival analysis with a proportional hazards model will then be used to compare time to dementia diagnosis in people who have had BSI compared with those who have not, adjusting for potential confounders.

2. Patients with Escherichia coli and Staphylococcus aureus BSI (n=24) and matched controls (n=12) will be prospectively recruited to a neuroimaging study where MRI scanning (high resolution T1-weighting and dynamic contrast-enhanced) will be performed shortly after and again 12-18 months after blood-stream infection to determine:
a. longitudinal rates of atrophy
b. blood-brain barrier dysfunction

Planned Impact

This project aims to develop a deeper understanding of the mechanisms by which infections adversely affect the brain. These findings are potentially of interest to several stakeholders:

1. Patients with blood-stream infections
Survivors of blood-stream infections and sepsis commonly report problems with memory and cognition. Research into this area has been identified as an important unmet patient need [UK Sepsis Trust - personal communication]. This study addresses a key question of whether blood-stream infections are associated with an increased risk of dementia and will identify patient groups most at risk. Furthermore, it will investigate the mechanisms by which this occurs. The longer-term aim of this project is ultimately to develop interventions that reduce the risk of infection associated encephalopathy at the time of infection and dementia risk subsequently.

2. Patients with other significant infections
This project uses blood-stream infections as a model of significant (i.e. resulting in hospitalisation) systemic inflammation cause by infection. Whilst patients with other types of infections are not studied in this project, it is likely that the underlying mechanisms of brain injury associated with infection are shared with other significant infections and sepsis. Therefore, the findings from this project and subsequent therapeutic interventions are likely to be generalisable to patients hospitalised with less well-defined infection syndromes.

3. Clinicians managing patients with blood-stream infections
The clinical management of blood-stream infections is inherently multidisciplinary. Therefore, by identifying therapeutic interventions that improve outcomes for survivors of blood-stream (and other significant) infections, this project will be of interest to microbiologists, infectious diseases physicians, acute and general physicians, GPs, neurologists and other healthcare professionals (e.g. nurses).

4. Researchers
This project has the potential to impact upon multiple areas of infection-related research including work on sepsis. The novel approach and findings from this work will benefit clinical researchers spanning disciplines of immunology, microbiology, infectious diseases as described in the 'Academic beneficiaries' section.

5. The general public
Sepsis is increasingly being recognised as a serious public health threat by the general public. Blood-stream infections are relatively common. In those who survive, their morbidity, in particular the cognitive morbidity, impacts not just patients but their families and carers. Given the ageing population and rising incidence of blood-stream infections the morbidity will increase without intervention. Furthermore, the findings from this study are likely to be relevant to a broad range of infections and will potentially lead to interventions to improve their outcomes potentially leading to less burden on society.

6. Policy makers
Morbidity following blood-stream infections is poorly understood. Although reporting of the more common organisms is mandated nationally, outcomes after hospitalisation have not been a focus of policy makers. This project will provide valuable data at a population level by determining the longer-term risk of dementia as well as mortality after hospitalisation. By highlighting the significant burden associated with these infections policy makers will be able to focus efforts not just on prevention but will be able to resource further studies to develop interventions to reduce the substantial human and economic costs.

Publications

10 25 50
 
Description CUBRIC collaboration 
Organisation Cardiff University
Department Brain Research Imaging Centre (CUBRIC)
Country United Kingdom 
Sector Academic/University 
PI Contribution i have led the collaboration with CUBRIC for the second part of my project. Here I plan to use the cutting edge facilities of CUBRIC to perform MRI scans on patients recovering from bloodstream infections to assess blood-brain barrier function.
Collaborator Contribution CUBRIC will be providing infrastructure support - MRI scanner, high-performance computing cluster for analysis and well as expertise in neuroimaging.
Impact Project significantly delayed by COVID
Start Year 2019
 
Description Collaboration with SAIL databank and Public Health Wales for a piece of episode work 
Organisation Public Health Wales NHS Trust
Country United Kingdom 
Sector Public 
PI Contribution I led the development of the collaboration. This will lead to the importation of bloodstream infection data into the SAIL databank to allow a detailed piece of work assessing the associations between bloodstream infections and non-infectious outcomes such as dementia, MI and CVA. Due to COVID the final data linkage has not yet taken place (PHW workers all redeployed to COVID as was I). This will likely be sorted in the next month or two as some sort of normality returns.
Collaborator Contribution PHW have data going back for many years and have extracted this. SAIL have the unique capability of then anonymously linking this data with many other forms of healthcare data from primary and secondary care to create a harmonised dataset that will permit detailed analysis of outcomes following bloodstream infections at a national level.
Impact This outcomes from this partnership have been delayed due to COVID
Start Year 2020
 
Description Collaboration with SAIL databank and Public Health Wales for a piece of episode work 
Organisation SAIL Databank
Country United Kingdom 
Sector Public 
PI Contribution I led the development of the collaboration. This will lead to the importation of bloodstream infection data into the SAIL databank to allow a detailed piece of work assessing the associations between bloodstream infections and non-infectious outcomes such as dementia, MI and CVA. Due to COVID the final data linkage has not yet taken place (PHW workers all redeployed to COVID as was I). This will likely be sorted in the next month or two as some sort of normality returns.
Collaborator Contribution PHW have data going back for many years and have extracted this. SAIL have the unique capability of then anonymously linking this data with many other forms of healthcare data from primary and secondary care to create a harmonised dataset that will permit detailed analysis of outcomes following bloodstream infections at a national level.
Impact This outcomes from this partnership have been delayed due to COVID
Start Year 2020