Research recommendations

The main priority for future research into symptomatic brain cavernoma is to conduct a definitive RCT comparing medical management and intervention (using neurosurgical resection or stereotactic radiosurgery) with medical management alone, to fully address the top priority for research into cavernoma.17 This pilot RCT has shown that a RCT of intervention can be done for this indication, and it has provided metrics that help to inform the design and feasibility of a definitive RCT.

In the CARE pilot RCT, we found that the majority of patients, who were either eligible but not approached or who were approached but were not uncertain, were reported to prefer medical management alone. Patients’ conversations reported by clinicians, QRI researchers and CAUK indicated that they would consider/prefer a non-invasive treatment. This confirms the need to address the fourth uncertainty identified by the James Lind Alliance Priority Setting Partnership for cavernoma, which was, ‘Could drugs targeted at cavernomas improve outcome for people with brain or spine cavernomas compared to no drug treatment?’17 Promising data exist for many drug treatments to prevent ICH from brain cavernomas,48,49 including propranolol,50 antithrombotic drugs51 and statins.52

Given these ongoing uncertainties about intervention and drugs for brain cavernoma, the most efficient way to resolve these therapeutic uncertainties seems to be an umbrella platform study,53 in which a definitive parallel-group RCT addressing intervention could be conducted using the same infrastructure as a multiarm, multiphase drug trial, such that a RCT would be available for all prevalent and incident patients with brain cavernoma other than people who had already undergone neurosurgical resection of a solitary symptomatic brain cavernoma.

Sample size required for a definitive randomised trial

We have used the feasibility outcomes and metrics from the CARE pilot RCT (see Tables 2 and 3) to estimate the sample sizes that might be required for a definitive RCT using the same primary clinical outcome as the pilot RCT. We assumed a recruitment duration of 5 years, an estimate of the recruitment rate (0.2 or 0.3 participants per site per month), a primary outcome event rate of 8% per year, annual withdrawal/dropout/death rate of 7% per year, a duration of follow-up of 10 years [as indicated in the NIHR HTA commissioning brief (Report Supplementary Material 1 and Report Supplementary Material 2)], a probability of falsely rejecting a true null hypothesis of 0.05 (using a two-sided test) and a power of the statistical test of 0.9. We chose to estimate the sample size and number of sites required by a definitive RCT (Table 8) using two estimates of recruitment rate (as above) and two estimates of a beneficial effect of intervention in reducing the primary clinical outcome (0.60 and 0.75) that are within the 95% CI of the estimate of effect in the pilot RCT and are similar to associations between intervention and a reduction in the primary clinical outcome in non-randomised observational cohort studies.23,26,54 Estimates assume that the maximum number of sites that open each month is 4, except for the HR of 0.75 and recruitment rate of 0.2 where this figure is taken to be 6 (as 4 would be insufficient).

Feasibility of a definitive randomised trial

The feasibility of a definitive main phase RCT is not only determined by the estimates above but also by the number of sites, countries and funding agencies that would be required to provide an adequate network to support recruitment over 5 years. The activation of 56–259 sites clearly mandates the involvement of additional countries, given that the UK and Ireland has 40 neurosurgical sites and only 28 (70%) of them took part in the CARE pilot trial. While we have good initial engagement and interest from professional networks (below), several additional countries may be needed for a definitive RCT. First, the number of neurosurgical sites in countries smaller than the UK will be lower. Second, the proportion of sites that can support RCTs in other countries may be lower than in the UK. Third, some large countries with good clinical research infrastructure (e.g. USA) have fee-for-service healthcare systems, which can complicate the conduct of pragmatic clinical trials of interventions already available in standard practice with levels of per-participant reimbursement that do not match the income generated for the care provider by the intervention under study. Fourth, the recruitment rate at international sites may not match the rate in a pilot phase in the lead country. Fifth, the letters of support we have received do not guarantee other countries’ inclusion, because they will require funding. Sixth, the cost of a definitive RCT will be large given that this pilot study cost over £1 M and will be most likely to be met by joint, collaborative funding from different agencies committed to commissioning research on the topic (thereby avoiding the ‘double jeopardy’ of approaching funding agencies independently). All of these challenges are not unique to brain cavernoma, but they have been surmounted for some neurosurgical RCTs in the past:55 in a review of 52 Cochrane reviews of neurosurgical interventions, seven focused on neurovascular surgery, among which UK-led, publicly funded international RCTs of ruptured intracranial aneurysm coiling versus clipping (ISAT) and surgical evacuation of acute intracerebral haemorrhage achieved sample sizes > 1000.

The feasibility of attaining the estimated recruitment rates will be determined by clinical communities’ awareness of, and reaction to, the results of the pilot RCT. Clinicians will not be aware of the results until April 2024, after which equipoise in the community can be reassessed. However, in anticipation of re-evaluating equipoise to determine feasibility, at the time of the application for this pilot RCT we obtained letters of support for a definitive RCT from professionals and patients:

1. Professionals
   1. European Organisation for Research and Treatment of Cancer
   2. International Paediatric Stroke Organisation
   3. Radiation versus Observation following surgical resection of Atypical Meningioma: a randomised controlled trial (the ROAM trial)
   4. UK Society of British Neurological Surgeons
   5. European Association of Neurosurgical Societies (EANS) vascular section
   6. EANS research committee
   7. European Stroke Organisation
   8. Italy Treat_CCM collaboration50
   9. Canadian HEmorrhagic Stroke trIals initiatiVE
   10. Australian neurology network
   11. Australia neurosurgery network
   12. Brazil neurology network
   13. France neurology network
   14. Germany neurosurgery network
   15. Germany neurology network
   16. Italy radiosurgery network
   17. USA neurosurgery network
   18. South Africa neurosurgery network
   19. Spain radiosurgery network
   20. Netherlands Collaboration Of New TReatments for Acute Stroke
2. Patients
   1. European Cavernoma Alliance
   2. Republic of Ireland
   3. Italy
   4. Sweden
   5. USA
   6. Spain
   7. France
   8. Germany
   9. UK