Some text in this manuscript has been reproduced from CARE Trial Pilot Collaboration.40 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: https://creativecommons.org/licenses/by/4.0/. The text below includes minor additions and formatting changes to the original text.
The main outputs that we included in the publication plan are:
- this synopsis
- the protocol1
- the CARE pilot trial report40
- the QRI substudy report.41
Cavernomas A Randomised Effectiveness (CARE) pilot trial
Full details about all outcomes are reported in the trial report,40 but we report a short synopsis of the main findings here and in Table 2. Between 27 September 2021 and 28 April 2023, we obtained Research Ethics Committee (REC) approval for 30 (75%) of 40 sites that were invited to take part, of which 28 (70%) completed site initiation visits and became active. We obtained an extension to the recruitment period in time due to slower-than-expected overall recruitment at a smaller number of sites than expected, where investigators screened 511 patients, of whom 322 (63%) were eligible. About 202 (63%) of eligible patients were approached for recruitment, of whom 96 had collective uncertainty with their neurosurgeon about whether to have intervention for a symptomatic brain cavernoma. A surge in recruitment during the last month of screening led to 72 (22%) of the eligible patients giving consent, who were randomly assigned at a median of 287 days (IQR 67–591) since most recent symptomatic presentation. Participants’ median age was 50.6 years (IQR 38.6–59.2), 68 (94%) were adults, 41 (57%) were female, 66 (92%) were White, 56 (78%) had prior ICH and 28 (39%) had prior epileptic seizure. The intended type of intervention before randomisation was neurosurgical resection for 19 (26%), stereotactic radiosurgery for 44 (61%) and no preference for 9 (13%). Baseline clinical and imaging data were complete for all participants. Randomisation assigned 36 participants to medical management and intervention (12 to neurosurgical resection and 24 to stereotactic radiosurgery) and 36 to medical management alone. Three (4%) participants withdrew, one was lost to follow-up and one declined face-to-face follow-up, leaving 67 (93%) retained at 6-month clinical follow-up. Sixty-one (91%) of participants with follow-up adhered to the assigned management strategy. The primary clinical outcome occurred in 2 of 33 participants assigned medical management and intervention [8.0% (95% CI 2.0 to 32.1) per year] and in 2 of 34 participants assigned medical management alone [7.5% (1.9–30.1) per year]. Investigators reported no deaths, no serious adverse events (SAEs), 1 protocol violation and 61 protocol deviations [deoxyribonucleic acid (DNA) sample not provided (n = 11); consent form errors (n = 10); delegation log errors (n = 10); baseline NIHSS score not recorded (n = 9); study visit done outside the time window specified in the protocol (n = 8); investigator training not completed (n = 5); 6-month visit not done face-to-face (n = 4); late DNA sample (n = 4); surgery done > 3 months after randomisation (n = 1)].
| Feasibility outcome | Metric(s) |
|---|---|
| What proportion of the collaborating sites take part and recruit participants to the CARE pilot trial? | 30/40 (75%) sites obtained REC approval. 28/40 (70%) sites were activated. 22/40 (55%) sites recruited ≥ 1 participant40 |
| Do investigators implement trial procedures correctly? | 61 protocol deviations. One protocol violation. Independent review confirmed definite certainty of brain cavernoma diagnosis in 70 (97%)40 |
| What proportion of screened patients is eligible? | 322 (63%) eligible of 511 screened40 |
| What proportions of eligible patients are approached and randomised (and why are eligible patients not approached or not randomised)? | 202 (63%) of 322 eligible patients were approached, and 72 (22%) were randomised following an extension to recruitment.40 Certainty about whether to have intervention was the main reason that patients were not approached or not randomised40,41 |
| What is the distribution of participants between neurosurgery and stereotactic radiosurgery? | The intended type of intervention that was pre-specified before randomisation for the 72 participants was neurosurgical resection for 19 (26%), stereotactic radiosurgery for 44 (61%) and no preference for 9 (13%)40 |
| Do participants adhere to the assigned management and follow-up? | Completeness of 6-month follow-up was 67/72 (93%).40 Adherence to assigned management was 61/67 (91%).40 Adherence to pre-specified type of intervention in participants assigned to intervention and medical management was 29/29 (100%)40 |
| How complete are baseline, imaging and outcome data? | Baseline clinical case report form complete 72/72 (100%).40 Baseline brain imaging received 72/72 (100%);40 6-month clinical follow-up complete 67/72 (93%);40 6-month brain imaging received 69/72 (96%)40 |
| What are the outcome event rates? (Rate of first outcome event during follow-up, quantified per participant per year) | First ICH/FND 7.8% (95% CI 2.9 to 20.7%). First ICH 3.8% (1.0%–15.1%). First FND 3.8% (1.0%–15.1%). Death not due to a primary outcome 0%40 |
| How do the baseline characteristics, outcome event rates and differences between treatment groups compare to observational data about outcomes during medical management alone or after intervention and medical management? | Baseline characteristics: demographics, clinical and imaging characteristics similar to systematic reviews/meta-analyses of untreated clinical course5 and after intervention.11,12 Outcome event rates: similar to rates described in the untreated clinical course4 and after intervention.11,12 Association between medical management and intervention vs. medical management alone with subsequent ICH ± FND in cohort studies: RR 0.6 (95% CI 0.2 to 1.7) over ≈5 years,23 RR 0.6 (0.1–2.6) over ≈4 years,26 HR 0.76 (0.4–1.4),40 RR 1.9 (0.6–6.0) over ≈6 years,24 aHR 3.6 (1.3–10.0) over 5 years10 and RR 7.8 (1.1–57.4) over ≈4 years22 |
| What estimates of effect size/variability should be used in the design of the CARE definitive randomised trial? | The observed unadjusted HR for the effect of medical management and intervention vs. medical management alone on the primary clinical outcome was 0.99 (95% CI 0.14 to 7.03)40 |
| What is the sample size required for a definitive trial to address the overall question over a 10-year follow-up? | Sample size estimates are provided in the research recommendations section |
| Can the CARE pilot trial data describe care pathways, linked to health states and outcomes, to develop a robust economic model to evaluate cost-effectiveness in a CARE definitive randomised trial? | Data about care pathways and outcomes were sufficient to estimate mean resource use and costs per participant as well as quality-of-life measurement for QALY estimation40 and to be used by a published Markov model20 |
QuinteT Recruitment Intervention
Full details about all findings, including quotes from investigators and patients, are to be found in the report of the QRI,41 but we provide a short synopsis of the findings here. The QRI substudy found several barriers: reluctance to offer randomisation to people recommended medical management alone within usual practice, reluctance to offer stereotactic radiosurgery (particularly for children or people with epilepsy), logistical challenges to review and recruit participants, especially during the COVID-19 pandemic, concerns about short follow-up in the trial and challenges of organising stereotactic radiosurgery. The substudy also identified several facilitators: investigators who were comfortable offering randomisation to people for whom medical management alone was usual practice, crucially with local multidisciplinary team support to do so and justifying the offer of stereotactic radiosurgery with reference to low risk of morbidity. The QuinteT and trial teams implemented multiple actions promoting recruitment at various stages in the pilot trial both before and during recruitment, including: balanced portrayal of the two trial arms in the participant information leaflet; emphasis on an inclusive approach to screening and training in optimising information provision about the trial at site initiation visits; videos to coach investigators in screening and recruitment discussions; tips and guidance documents for investigators; and investigator meetings.41
Success metrics proposed to the National Institute for Health and Care Research
Table 3 shows our findings in relation to the metrics we proposed to the NIHR in the contract; we fulfilled the majority of these metrics, and the one metric that we failed (retention > 95%) came close (93%) and would have been achieved had two more of the 72 participants been followed up at 6 months.
| Success metric | Findings | Achieved? |
|---|---|---|
| At least 30 sites in the UK and Ireland collaborate | 30 sites (28 in England, Wales and Scotland; 2 in Ireland) obtained REC approval. 28 sites (in England, Wales and Scotland) were activated40 | Yes |
| Project delivered according to the major milestones identified in the timetable | We delivered the project according to the project milestones with a 5-month extension (awarded due to delays attributable to the impact of COVID-19 pandemic)40 | Yes |
| Recruitment to within 10% of target | The trial recruited 72 participants at 28 sites over 23 months (mean recruitment rate 0.2 ± 0.25 participants per site per month), which exceeded the recruitment target of 60 participants at 40 sites over 18 months (recruitment rate 0.11 participants per site per month)40 | Yes |
| Brain cavernoma radiographic diagnosis confirmed by expert neuroradiologist review in > 95% of patients recruited | Expert review of diagnostic brain imaging confirmed definite certainty of brain cavernoma diagnosis in 70 (97%) and probable certainty in two (3%) participants40 | Yes |
| Retention of > 95% of participants at 6 months | 67 (93%) participants were retained at 6-month clinical follow-up40 | No |
| < 10% treatment group switches or lost to follow-up | 61 (91%) of 67 retained participants adhered to the assigned management strategy40 | Yes |
| QRI is associated with an improvement in recruitment | Visual interpretation of the recruitment graph indicates that the recruitment rate increased over time in association with the conduct of the QRI,40,41 but other potential confounders (e.g. the number of sites) changed at the same time40 | Maybe |
| Definitive RCT appears feasible and affordable | We have estimated a range of potential sample sizes for a definitive RCT, which would require a multicentre, international network to recruit sufficient participants over 5 years. We have had expressions of interest from a range of potential international partners | Maybe |