MR1447 - Long-term follow-up of Asymptomatic Carotid Surgery Trial (ACST-1)

This Data Sharing Agreement permits the retention of the data provided under previous iterations of this Agreement for an interim period. This is a pragmatic approach to provide an active Agreement whilst enabling The University of Oxford to complete the necessary actions to enable a subsequent application to extend the Agreement meeting all applicable data sharing standards as published in NHS Digital’s website (see: https://digital.nhs.uk/services/data-access-request-service-dars/dars-guidance).

The University of Oxford wishes to retain data for the purpose of a study involving the long-term follow-up of the Asymptomatic Carotid Surgery Trial (ACST-1) cohort. This is to allow publication of analysis on long-term stroke, and possibly studies that have become ‘of interest’ following the publication of ACST-2 results (Lancet 2021).

Stroke is the third leading cause of disability and premature death worldwide. Around 20% of all strokes are caused by narrowing of carotid arteries, when fatty deposits break off from such lesions and lodge in the brain, thereby causing a stroke and over 80% of these carotid related strokes occur without warning. ACST-1 showed that preventative surgery to remove carotid narrowing (carotid endarterectomy) halves the risk of stroke up to 10-years following a successful procedure and these operations are performed widely worldwide (around 200,000 per year in US and Europe). There is a strong association between carotid stenosis and dementia, but (unlike the link between carotid disease and stroke) it is not clear if this association is causal, and whether carotid surgery could also prevent cognitive decline. Determination of causality requires randomised evidence, and this is the purpose of this study.

The legal basis for processing falls under GDPR Article 6(1)( e) and GDPR Article 9(2)(j). Processing and dissemination meet public interest criteria as defined by the ICO for good decision-making by public bodies and securing the best use of public resources.

ACST-1 (ISRCTN26156392), randomly allocated 3120 participants with asymptomatic narrowing of the carotid artery to carotid surgery (endarterectomy) plus best medical therapy or to best medical therapy alone. Procedural complications (i.e. those occurring within 1 month) were recorded by surgeons collaborating in the trial, and thereafter participants informed the trial coordinators of any late strokes by means of an annual postal questionnaire, the last of which was mailed in 2008. Cause specific mortality was also provided by ONS for the 1069 UK participants. The relationship between carotid disease and dementia was not well recognised when ACST-1 was designed in the early 1990s, nor was dementia the major public health priority that it now is. Accordingly, ACST-1 did not routinely collect information on cognitive decline. However, this trial provides a unique opportunity to reliably assess the effect of carotid surgery on dementia. This question is also being assessed (at considerable expense) by a similar trial currently underway in North America (CREST-2), but meaningful results may only emerge in around 10 years’ time.

The University of Oxford is the sole Data Controller and the sole Data Processor. This study will be conducted by the MRC Population Health Research Unit, Clinical Trial Service Unit (CTSU), Nuffield Department of Population Health, which is based at the University of Oxford. The aim of the Unit is to generate and disseminate reliable evidence from observational epidemiology and from randomised trials that leads to practicable methods of avoiding premature death and disability. Prevention of premature vascular death and morbidity is a major focus of CTSU’s work, and the effects of drugs such as statins and aspirin on dementia are currently being investigated (led by members of CTSU). Consequently, the proposed study fits within an established research program, and the required specialist skills and knowledge exist at CTSU.

Whilst CTSU have data on stroke rates at up to 10-years post-surgery, CTSU did not collect information on incident dementia during ACST-1 follow-up which ended in 2008, and CTSU have no information on any health outcomes after 2008. Indirect follow-up via routinely collected health records is necessary to assess rates of dementia in this cohort of participants thereby allowing a complete and unbiased follow-up of the UK ACST-1 participants.

The data held prior to 2008 does not contain information on incident dementia. The CTSU seeks data on dementia from before this date as well as after to answer an additional question on whether patients who had carotid surgery (endarterectomy) in the original cohort also had a lower risk of contracting dementia.

Dementia is an insidious process which takes a number of years (usually decades) to develop which is why if only a short time was requested CTSU would miss potential dementia and stroke diagnoses. The risk of stroke and dementia increases with age which is why this is such an important study in a high-risk elderly cohort. CTSU know that the 5 and 10 year results of ACST-1 trial highlighted the benefit of carotid endarterectomy however CTSU do not know the longer term and potentially enhanced benefits seen at >20 years post-intervention. Data on incident dementia is required going back to as early as possible for this reason.

The study is not aware of any moral or ethical issues raised by the proposed dissemination of NHS Digital data to University of Oxford, nor the dissemination of the aggregated study results to the wider public. The study is unaware of any risk of potential harm to the public by the dissemination.

ACST-1 Long-term follow-up study is funded by an Alzheimer’s Society project grant and also by core funding from the Clinical Trial Service Unit (CTSU), University of Oxford.

The data supplied to NHS Digital by CTSU for each participant was: unique ACST-1 trial Identifier (ID); first and last name; date of birth, and NHS number. The data was was sent in an encrypted format via the Oxford secure data transfer system.

The data received back from NHS Digital included the ACST-1 trial ID, date of birth (as provided by CTSU), gender, and outcomes of potential relevance (e.g. strokes and dementia) from the following datasets: HES Admitted Patient Care, HES Outpatients, and A&E, Mental Health, Civil Registration Mortality (including Date of Death and Cause of Death, and National Dementia and Anti-psychotic Prescribing Audit (DAP). The data was securely transferred to the secure database at Clinical Trial Service Unit (CTSU), Richard Doll Building, University of Oxford. Upon receipt, these data were entered into the existing ACST-1 trial database, and hence will remain identifiable. However, access to this database is tightly restricted, and any outputs are aggregated and entirely non-identifiable.

Date of birth was required back from NHS Digital in order apply a two-fold validation of the individual records by linking participants in the original trial dataset with their health recorded in the hospital episode statistics. This is due to policy at CTSU to employ rigorous data-checking procedures for all imported data. By doing this, CTSU can have higher confidence that matching of data is accurate.

The ACST-1 trial has a study specific database that is password protected and held securely on one University drive only. The ACST-1 database contains detail from the original ASCT-1 study where members are archived according to the original paper consent forms and records. These records include full name and date of birth, but these are the only identifiable data contained within the ACST-1 database. The identifiable HES data is stored in an encrypted TrueCrypt container where all such access was granted on the instruction of the Information Asset Owner for ACST-1. The database with imported NHS Digital information is modified and only Trial ID and age was used in the database for analysis. Pseudonymised data (with exception to date of death) was available for analysis. Access is routinely reviewed and revoked when the team member leaves ACST-1. Only personnel substantively employed by the University of Oxford and are involved in this long-term follow-up study for ACST-1 study (processing and analysing data) will have access to this data.

Cause specific mortality is required for three reasons:

1) Since the primary outcomes will be time-to-event analyses, the date of death is required to determine the population at risk for stroke and dementia (censoring);

2) Detect incident stroke;

3) Detect incident dementia (which is frequently recorded in death certificates as the underlying cause of death or as a contributing factor)

The National Dementia and Anti-psychotic Prescribing Audit (DAP) will identify UK ACST-1 participants who were prescribed medications exclusively used in the management of dementia, thereby serving as a dementia-diagnosis equivalent.

The data has been combined with existing information on rates of stroke, other vascular disease and death (obtained previously up to 2008) and used to compare rates of dementia (and also stroke) amongst participants originally allocated to undergo carotid surgery versus those managed with medical therapy alone (i.e. intention-to-treat analyses).

All organisations party to this agreement must comply with the Data Sharing Framework Contract requirements, including those regarding the use (and purposes of that use) by “Personnel” (as defined within the Data Sharing Framework Contract ie: employees, agents and contractors of the Data Recipient who may have access to that data).

No data will be shared with collaborators in the UK or Sweden including NHS Scotland, University of Uppsala Sweden, University Lund Sweden or with any other 3rd parties.

The Data will only be used for the purposes described in this agreement.

There will be no data linkage undertaken with NHS Digital data provided under this agreement that is not already noted in the agreement.

Data will only be accessed and processed by substantive employees of the University of Oxford and will not be accessed or processed by any other third parties not mentioned in this agreement.

All outputs will consist of aggregate data only with small numbers suppressed in line with the HES analysis guide.

Results of the analyses were presented at relevant international scientific meetings 2019-2020. The results of this analysis were posted on the ACST-1 section of website www.acst-2.org in a similar manner to multiple previous ACST-1-related publications and were also available through the ISRCTN registry. The publication of the results of analyses was in 2019-2020.

If ACST-1 can reliably demonstrate that carotid intervention (carotid endarterectomy or carotid stenting) in middle age leads to a reduction in the risk of dementia, stroke or death in the longer term this may increase the number of carotid interventions performed and help patients make better decisions about their treatment. The outcome of the research will inform clinicians and patients about potential preventative treatment pathways. Any change in medical practice will need to be discussed with advisory boards (eg, NICE) but patients will be informed of the outcome in discussions with their doctor. CTSU will be taking the Alzheimer’s Society advice with regard to disseminating this information into the public forum.

All results will be presented at international scientific meetings and published in a prominent peer-reviewed medical journals. The results of the research will be disseminated in leading peer-reviewed, high impact-factor journals and conferences including the American Heart Association, UK Stroke Forum, European Society of Cardiology, European Society for Vascular Surgery and the European Stroke Organisation Conference.

Preliminary dementia diagnosis and methodology data was presented at the Alzheimer's Society Annual Meeting in 22-23 May 2018

Long-term (15 year) stroke risk/outcome for ACST-1; long-term persistence of the early benefits presented at the European Cardiology Congress, Munich August 2018

The paper 'Effect of carotid endarterectomy on 20-year incidence of recorded dementia: a randomised trial' has been accepted for publication in the European Journal of Vascular & Endovascular Surgery' for 2022.

The ACST-1 trial showed that early carotid surgery prevents future fatal and disabling strokes for at least 10 years. This finding changed clinical practice worldwide, with rates of asymptomatic carotid intervention rising sharply in the UK and Europe. However, due to improvements in medical therapy (chiefly better lipid-lowering drugs) there is renewed uncertainty about the benefits of preventative carotid surgery and rates of intervention are falling (particularly in the UK). If this study shows an additional benefit of carotid surgery on dementia risk (NB: Statins do not protect against dementia in several very large, randomised trials), then rates of carotid intervention may rise again, thereby preventing large numbers of strokes and also helping to prevent dementia. Rates may rise once advisory boards have been notified. This is common practice in any change of medical practice. E.g. NICE, and the Royal College of Surgeons will need to be involved to advise on the best treatment dependent on results.

CTSU know that at 5 and 10 years, carotid endarterectomy reduces stroke (Lancet 2004, 2010). However, CTSU do not know whether it still reduces stroke rates in the longer-term (e.g. median 22 years). In addition, no other large randomized controlled trial has investigated whether this type of surgery reduces the rate of dementia with results within the next year. No other surgical study has looked at dementia. This is the first and CTSU have a large cohort. There are other surgical studies looking at dementia, but their results will not be available for a number of years. The ACST-1 trial 5, and 10 year results (Lancet 2004, 2010) showed that surgery reduced the risk of stroke at 5 and 10 years after the operation substantially and changed medical practice worldwide. CTSU anticipate that the results of this long-term study will have a similar impact.

Dementia develops over many years, and intervention may take a long time to have an effect so long-term follow-up is vital. The ACST-1 trial has >15 years of follow up and is therefore uniquely placed to help answer the question of whether increased rates of carotid intervention prevent large numbers of strokes and help prevent dementia. It may be able to detect an effect of carotid intervention on dementia more reliably than any previous study. The study is funded by a grant from the Alzheimer’s Society who supports this research question as part of their emphasis on dementia prevention and also by the University of Oxford as sponsor*, the Clinical Trials Service Unit and the Nuffield Dept. of Population Health.

*'Sponsor' in this context means the legal establishment which:

• takes legal responsibility for initiation and management of the study;

• provides insurance for the study;

• is required for all clinical research not just trials

• is not necessarily the funder. Many funding bodies are unable to provide Sponsorship.

The delays to the analysis are a direct consequence of COVID, with extreme clinical pressure on the Co Investigator and WHO COVID work for the statistical PIs named in the study's protocol.

Details of yielded benefits will be updated in a subsequent Agreement.