Background

Chronic subdural haematoma (CSDH) is an ‘old’ collection of blood and blood breakdown products in the subdural space. It is radiologically defined as a predominantly hypodense or isodense collection in the subdural space along the cerebral convexity on computed tomography (CT). It is especially common in older patients and in the UK, 5000 people aged > 65 years are diagnosed with a CSDH each year. It can happen following only a minor injury to the head or even in the absence of a known trauma [1]. Symptoms that can be attributed to a CSDH include headache, gait disturbance, falls, cognitive decline, focal neurological deficit, speech disturbance, decreased consciousness and seizures.

Patients with severe symptoms usually undergo an operation to evacuate the CSDH; while around 80% of patients recover well, around 10–20% experience recurrence of the CSDH requiring further surgery [1, 2]. Evidence from a previous CSDH trial looking at subdural drains demonstrated that a reduction in recurrence resulted in reduced mortality and rate of poor functional outcome at six months [1]. A considerable body of evidence suggests that administration of steroids could reduce CSDH recurrence and even the rate of primary surgical intervention [2,3,4,5]. This, in turn, might be expected to reduce mortality and morbidity and improve long-term functional outcome in patients with CSDH. While the mechanism of action of steroids in CSDH is not entirely understood, recent research suggests that inflammation may be responsible for driving the continued growth of CSDH [6,7,8,9]; therefore, steroids may help overcome this.

The Dex-CSDH trial (DEXamethasone in Chronic SubDural Haematoma) is a multi-centre, pragmatic, clinical phase III, randomised, double-blind, placebo-controlled trial of dexamethasone for up to two weeks in patients diagnosed with CSDH. Dexamethasone is one of the most potent synthetic analogues of the naturally occurring glucocorticoid hydrocortisone and has practically no water- and salt-retaining properties, so is suitable for use in patients with cardiac failure or hypertension [10]. The earliest application of steroids in neurosurgery was for patients with brain tumours and surrounding oedema, where 4 mg four times a day was established as the dose with maximum effect [11]. This dosing, with subsequent gradual weaning, continues to be used in neuro-oncology and a two-week course of dexamethasone was considered likely to provide the best balance in terms of clinical efficacy and risks in this study [12]. The dose and duration are also reflective of other studies in the field [13].

The potential impact of this trial is significant, as the results will determine whether steroids should be prescribed routinely for patients with symptomatic CSDH. If steroids are found to be effective, an impact on the speed of recovery and functional outcome of patients is expected. This will be measured by the primary outcome, the modified Rankin Scale (mRS) at six months. Additionally, this could reduce the rate of surgical interventions required, length of hospital stay, discharge destination and adverse events (AEs). As well as the impact on clinical outcome, there are health economic considerations that will be addressed by the trial.

Trial rationale

We hypothesise that a two-week course of dexamethasone can improve the six-month functional outcome of patients with symptomatic CSDH by reducing the rate of CSDH-related surgical interventions and the recurrence rate.

Trial objectives

Primary objective

To detect an 8% absolute difference in the rate of favourable outcome at six months between the two arms.

Secondary objectives

  • Compare the long-term clinical effectiveness of dexamethasone versus placebo (six-month follow-up period)

  • Compare the AEs and complications between the two arms

  • Undertake a detailed economic evaluation between the two arms

Exploratory (mechanistic) objectives

  • Assess the biological action of dexamethasone with CSDH fluid and blood analysis

  • Assess the role of dexamethasone in cerebral perfusion and swelling in CSDH

Methods

Study setting

All study sites are in the UK. Patients are admitted to their local neurosurgical unit (NSU) following diagnosis of CSDH on CT. Local clinical neurosurgical teams review patients upon admission to the NSU and will assess eligibility for the Dex-CSDH trial. The decision for surgery or active monitoring is made on an individual patient basis by the admitting clinical team in conjunction with the patient and their families. This will not be affected by trial involvement, with both surgical and conservatively managed patients eligible for trial recruitment.

Eligibility criteria

Screening of patients to determine eligibility for participation in the trial will be undertaken by the neurosurgical team upon admission to the NSU according to the following criteria.

Inclusion criteria

  • Adult patients (aged ≥ 18 years)

  • Symptomatic CSDH confirmed on cranial imaging (e.g. CT/magnetic resonance imaging [MRI]), predominantly hypodense or isodense crescentic collection along the cerebral convexity on CT

  • Informed consent or Independent healthcare profession (IHP) authorisation

Exclusion criteria

  • Patients with conditions where steroids are clearly contraindicated

  • Patients who are on (or within one month of) regular oral or intravenous glucocorticoid steroids

  • Previous enrolment in this trial for a prior episode

  • Time interval from the time of the admission to the NSU to the first dose of the investigational medicinal product (IMP) > 72 h

  • CSDH in the presence of a cerebrospinal fluid shunt

  • Severe lactose intolerance or any known hypersensitivity to dexamethasone or other IMP excipients

  • Patients with a previous history of psychotic disorders

  • Unwillingness to take products containing gelatin

  • Concurrent enrolment in any other trial of an IMP

  • Biochemical sub-study only: active malignancy or currently receiving immunosuppressive drug therapy

  • MRI sub-study only: renal dysfunction, pacemaker or metal implants

Interventions

The trial aims to run in parallel to standard clinical care. The only difference between the trial pathway and the standard NHS pathway is the addition of a two-week tapering course of either dexamethasone or placebo (as per Table 1).

Table 1 Trial dosing regimen
Full size table

The trial treatment can be delivered orally or by nasogastric tube. In special circumstances (such as patients who are nil by mouth for surgery), where study medication is missed at lunchtime, that day’s dose(s) may be taken later as long as it is on the same day. Otherwise, in the event of missing a dose of medication, these can be taken when remembered, but only up to the time of the next planned dose on the same day. No dose modifications are permitted within this trial. The trial is being carried out under a Clinical Trial Authorisation (CTA); for a list of known drug reactions and interaction with other therapies, see Appendix 1.

Irrespective of whether an operation is undertaken, patients will complete the two-week course of trial medication. Patients may be discharged or transferred to a local hospital before the completion of the two-week course; in this case, letters will be provided to the patient and medical and pharmacy teams at the local hospital along with the remaining trial medication to ensure that the course is completed. The exception to this will be in the event of a patient receiving study drug via the nasogastric route, where it will be stopped at discharge/ transfer if this is the case.

Trial teams will ensure compliance with treatment is documented, using source data which should include the inpatient medication administration record and the trial medication diary, as well as performing physical capsule counts during inpatient treatment where possible. Please refer to the Dex-CSDH IMP Handling Manual for further information.

Any concomitant therapy clinically required will be permitted, including gastroprotection as per local policy. A list of contraindicated concomitant therapies to be avoided during the trial is detailed in sections 4.3 and 4.5 of the current SmPC for dexamethasone [14]. Only concomitant therapies of interest will be recorded on the concomitant medication log in the case report form (CRF), including: gastroprotection; anti-diabetic medication; and single (intraoperative) dose of dexamethasone.

Trial outcome measures

Primary outcome measure

mRS at six months after randomisation (Table 2). This scale was selected as it is a core instrument for measuring the degree of disability or dependence in daily activities of living and has previously been used in CSDH studies and stroke research, which affects a similar patient demographic [1, 15, 16].

Table 2 Modified Rankin Scale (mRS)
Full size table

Secondary outcome measures

  1. 1.

    Number of CSDH-related surgical interventions undertaken during the index admission

  2. 2.

    Number of CSDH-related surgical interventions undertaken during subsequent admissions in the follow-up period

  3. 3.

    Glasgow Coma Scale (GCS) at discharge from NSU and at six months

  4. 4.

    mRS score at discharge from NSU and at three months

  5. 5.

    Barthel Index at discharge from NSU, three months and six months

  6. 6.

    Mortality (at 30 days and six months)

  7. 7.

    EuroQOL (Quality of life) EQ-5D at discharge from NSU, three months and six months.

  8. 8.

    Length of stay in NSU

  9. 9.

    Discharge destination from NSU

  10. 10.

    Length of stay in secondary care

  11. 11.

    Health-economic analysis

  12. 12.

    AEs

Economic evaluation

An economic analysis will be conducted alongside the trial. Costs will be estimated from the viewpoint of the NHS and personal social services. Resources associated with provision of dexamethasone will thereby be monitored along with any surgical operation(s) to evacuate the CSDH, length of stay in NSU and any further hospital admissions /surgical procedures, e.g. for recurrence of the CSDH. Additionally, the level of informal care will also be monitored to estimate the opportunity cost for family, friends, carers and patients.

Exploratory (mechanistic) outcome measures

To assess the mechanism of action of dexamethasone, we will be collecting CSDH fluid and blood samples on selected patients who undergo surgery as part of their standard clinical care. Analysis of inflammatory biomarkers will be performed on blood and CSDH fluid and compared between the dexamethasone and placebo patients. Transcranial Doppler (TCD) and MRI may also be used in a sub-set of patients to measure cerebral blood flow patterns and assess whether this can be used to predict recovery and recurrence from CSDH.

Participant timeline: trial assessments and schedule

All patients will have a medical history taken and a clinical examination as part of the routine standard of care, including: past medical history; injury-related events; neurological status; imaging (modality, date of examination and original images at selected sites); and routine lab results. Additional data will be collected on the exploratory outcomes (CSDH fluid, blood, TCD and MRI) if applicable, at the sponsor site only. Full details are available in the DEX-CSDH Laboratory and Imaging Manuals. Figure 1 shows a full schedule of trial assessments as per SPIRIT guidelines.

Fig. 1
figure 1

Schedule of assessments. * = only collected in patients recruited to sub-study in coordinating centre. Ax within 72 h of admission to NSU, AE adverse event, D day, D/C discharge (or death if sooner), EQ-5D European Quality of life-5 dimensions, IMP investigational medicinal product, I-O intraoperative, mon months, mRS modified Rankin Scale

Full size image

Sample size

A sample size of 750 patients was determined with a power in the range of 81–92% and a two-sided significance of 5%, allowing for 15% missing data.

Recruitment

The study commences with an internal pilot, stage 1 (feasibility study) to ensure 100 patients can be recruited by a limited number of centres within 12 months. Following successful completion of this, stage 2 (substantive study) will take place. The recruitment rate has been estimated at two patients per site per month. On the basis of Hospital Episode Statistics (HES) and data from the national CSDH audit [17], approximately 60–80 patients with a CSDH are admitted in a medium-sized NSU each year. Hence, the estimated recruitment rate is feasible. Patients will be monitored while in the acute NSU and followed up for a period of six months after recruitment.

Treatment assignment, randomisation and blinding

Patients will be randomly assigned to either the control or intervention group with a 1:1 allocation as per a computer-generated randomisation schedule stratified by site using permuted blocks of random sizes. An interactive web-based response system (IWRS) will be used for allocating treatment packs to individual patients once confirmation that the inclusion criteria have been met has been confirmed.

Placebo is a capsule, visually indistinguishable from the active treatment and containing inactive excipients only. Dexamethasone capsules will consist of over-encapsulated dexamethasone 2 mg tablets. A proprietary brand will be used. The study drug will be supplied in individually numbered patient bottles. Capsules and packaging for both active and placebo arms will be identical in appearance at the point of issue to patients.

It is estimated that < 10% of eligible patients will have (or develop during the trial) swallowing difficulties, making oral IMP administration difficult or impossible/unsafe. In such cases, the blinded capsules may be opened at the point of administration by ward nursing staff and the contents dispersed in water, for administration either via oral route or a nasogastric tube. The administering nurse and potentially the trial patient will no longer be blinded, because the active dexamethasone is in tablet form that has been over-encapsulated and the placebo will be in powder form. To maintain blinding of the neurosurgeons, the presence of tablets being inside the opened capsule should not be documented in the medical notes.

Every effort must be made to maintain patient blinding when NG administration is used, by the patient not seeing the capsules being opened. Should, despite these efforts, the patient discover their treatment, they should be asked to not disclose their treatment allocation to any of the other medical personnel they interact with, e.g. surgeons, etc. The research staff and outcome assessors will remain blinded.

There are also clinical aspects that could potentially unblind trial team members to treatments allocated. Patients receiving dexamethasone will be more likely to have higher blood glucose levels compared to those receiving placebo. This may provide an indication but not proof that a patient is in the active arm. Concealment of glucose measurements will be difficult as clinical action may be required.

Any decision about surgery is made based on the severity of symptoms and/or progression of symptoms. Therefore, in cases where the IMP has been started before any neurosurgical intervention, a hint that the patient is in the active arm would have little influence on decisions about operative or non-operative management. Overall, we anticipate this occurring in such a small number of patients that the risk of bias is negligible and will not affect the overall findings of the study.

Emergency unblinding

Emergency unblinding will be managed according to the emergency unblinding procedure using the IWRS. Emergency unblinding requested by the patient’s clinical team will only occur in exceptional circumstances (e.g. need to treat a serious adverse event [SAE]) when knowledge of the actual treatment is essential for further management of the patient.

Patient withdrawal

Each patient has the right to discontinue their participation in the trial at any time. If an unconscious patient regains capacity and makes a request to be withdrawn from the trial then this is accepted. Incapacitated patients may also be withdrawn from the trial if the consultee requests withdrawal. In addition, the investigator may withdraw the patient from their allocated treatment arm if, subsequent to randomisation, a clinical reason for not providing the drug treatment is discovered.

As the trial will be analysed on an intention to treat basis, any data collected will remain in the trial and the patient will continue to be followed up unless consent is withdrawn. Patients who have been withdrawn from the trial will not be replaced as the power calculation for the trial allows for a 15% loss to follow-up. All discontinuations and withdrawals will be documented. If a patient wishes to discontinue, anonymised data collected up until that point will be included in the analysis.

Consent, enrolment and data collection

All patients who have been admitted to the NSU with a confirmed CSDH may be screened for eligibility. Screening will be carried out by a member of the clinical team and a log kept. Consent must be taken before study randomisation and study drug administration.

Where potential patients fulfil the eligibility criteria, they will be approached by a member of the research team who will provide the patient information sheet and clarify any information from the patient/relatives which may preclude recruitment. At Cambridge only, patients will also be screened for eligibility for the exploratory sub-studies. If they are eligible, they will be given an additional page in the patient information and consent sheet so that they can consider if they would like to take part in any of the additional sub-studies. If they do not wish to take part in these, it will not affect their recruitment to the main trial.

Wherever possible, informed consent will be obtained from the patient. However, due to the nature of the condition, this may not always be possible. If lacking capacity, patients with CSDH can still be enrolled in the trial if consent is obtained from:

  1. i.

    the patient’s legal representative (if available in the hospital);

  2. ii.

    IHP consent (if a patient’s legal representative is not available in the hospital) - this can be completed by someone who is not connected with the conduct of the trial, specifically:

    1. a.

      the sponsor of the trial;

    2. b.

      a person employed or engaged by or acting under the arrangements with the sponsor, and who undertakes activities connected with the management of the trial;

    3. c.

      an investigator of the trial; or

    4. d.

      a healthcare professional who is a member of the investigators’ team for the purposes of the trial.

Patients who regain capacity will be informed about the clinical trial and consent to continue will be sought during their in-patient stay and if still lacking capacity on discharge, at their six-month clinical follow-up appointment (if attended). If at any stage either the legal representative or the patient chose to withhold consent, then the patient will be withdrawn from the trial.

All enrolled patients will be offered an optional study wrist band, to be applied before their first dose of medication and worn while they are an inpatient. This highlights that they are taking part in a blinded study and helps the patient and nursing staff be aware of the study at all times and to reduce the risk of open label ward dexamethasone stocks being used in error.

Patients will be monitored as per routine clinical practice in the NSU until discharge and thereafter at approximately three and six months to assess clinical outcome. Follow-up will be by postal questionnaire. However, if after two weeks the questionnaire has not been returned, patients will be followed up by telephone. If after a further four weeks there is no response, then the patient will be deemed as lost to follow up. Where patients attend for a routine clinical follow-up, they will be reviewed by a blinded assessor.

Data management

A final trial report will be written for publication and trial results will be presented internationally at meetings. All data will be entered into a CRF, which will be anonymised. The CRF will be accessible to trial coordinators, data managers, the investigators, Clinical Trial Monitors, Auditors and Inspectors as required. All CRF pages will be completed in a Good Clinical Practice (GCP)-compliant manner. All investigators and trial site staff involved in this trial must comply with UK Data Protection requirements and Trust Policy with regards to the collection, storage, processing and disclosure of personal information.

Statistical methods

Analysis will be performed on an ‘intention-to-treat’ basis. The primary endpoint is the mRS at six months which is then dichotomised to favourable (0–3) versus unfavourable (4–6). The primary analysis will estimate the absolute difference between the two treatment arms in the proportions achieving a favourable outcome. A normal approximation will be used to produce 95% confidence interval and a two-sided p value testing the null hypothesis of no difference. Secondary analysis will include a proportional odds logistic regression of the mRS score adjusting for baseline covariates (age, GCS).

Assuming a favourable outcome rate of 80–85% in the control group, an 8% increase in the rate of favourable outcome (mRS 0–3) at six months is a plausible and clinically important treatment effect [1]. Using a two-sided test at the 5% significance level, a sample size of 750 patients (allowing for a 15% loss to follow-up) will enable us to detect this 8% absolute difference in the rate of favourable outcome with a power of 81–92%.

Further secondary endpoints will be summarised using appropriate techniques according to whether the variable is binary, categorical, continuous or time-to-event. Categorical and binary variables will be summarised using bar charts, frequency tables and comparisons made using logistic regression. Continuous variables will be summarised, broken down by treatment arm, using Box plots, mean, median, SD, max, min and compared using linear regression. Time-to-event variables will be summarised using Kaplan–Meier plots, and compared using the log-rank test.

Economic analysis

Appropriate unit costs will be assigned to each item of the aforementioned items of resource use (see trial outcome measures) using a standard price year. The mean incremental cost for those allocated to dexamethasone compared to placebo intervention over the six-month trial period will then be estimated, from both an NHS and personal social services perspective, and also with the addition of informal care costs. Assuming dominance does not occur (where one option is estimated to be more effective and less costly that the other option), the incremental cost-effectiveness ratio associated with dexamethasone will be estimated and assessed in relation to a range of cost-effectiveness thresholds, e.g. £20,000–30,000 per quality-adjusted life years (QALY) is recommended by National Institute for Health and Care Excellence (NICE) [18]. The associated level of uncertainty will also be characterised by estimating cost-effectiveness acceptability curves [19]. Additionally, sensitivity analysis will also be undertaken to assess the robustness of conclusions to change in key assumptions. In line with the outcome analysis, all analyses will initially be conducted on an intention-to-treat basis.

Interim analysis

An interim analysis, blinded to all except the study’s Independent Data Monitoring and Ethics Committee (IDMEC), will be performed after an appropriate number of patients have reached the six-month follow-up, to confirm the final sample size. The Trial Steering Committee (TSC), IDMEC and statistical team will agree jointly on the most appropriate timing of this interim analysis, taking into account the case mix and parameters the IDMEC wishes to estimate. If the sample size needs to be revised, we are able to incorporate the uncertainty in absolute favourable outcomes rates (80–85%) in order to achieve an acceptable conditional power as determined by the IDMEC. If sample size adjustment is necessary, the final analysis will adjust for the inflated type 1 error rate. The primary purpose of the internal pilot (first 100 patients) is to assess recruitment rates rather than to make sample size adjustments.

There are no defined criteria for the premature discontinuation of the trial. However, the IDMEC and TSC will make recommendations on the discontinuation of the trial following review of the ongoing patient safety and efficacy data presented at regular scheduled meetings. For the primary analysis, missing data will be assumed to be missing at random. A sensitivity analysis will be carried out by performing a complete case analysis. As the relevant covariates need to be recorded before the patient can be randomised, we aim to have minimal missing baseline data. There is also an excellent track record for UK-led neurosurgical studies in achieving extremely high rates for follow-up [20,21,22].

The end of the trial is the date that the last expected six-month follow-up questionnaire is completed for the last-recruited trial patient.

Trial monitoring and safety

The TSC will provide overall supervision with respect to the conduct of the trial and be independently chaired by Professor Anthony Bell (St George’s, University of London, London, UK). The ethical and safety aspects of the trial will be overseen by an IDMEC, which will be chaired by Professor Martin Smith (The National Hospital for Neurology and Neurosurgery, London, UK).

The competent authority, the Medicines and Healthcare Products Regulatory Agency (MHRA), provided clinical trials authorisation before trial commencement. The protocol and trial conduct will comply with the Medicines for Human Use (Clinical Trials) Regulations 2004 and any relevant amendments. Development Safety Update Reports and Annual Safety Reports are submitted to the MHRA in accordance with UK requirements. It is the Chief Investigators responsibility to produce the annual reports as required.

Due to the patient demography and the clinical condition of CSDH, there may be many AEs throughout the initial admission. All patients are regularly monitored either in the intensive care environment or on the neurosurgical wards, but it is not practicable to record all AEs. Therefore, only AEs of special interest (AESIs) and SAEs will be reported. Some SAEs will be classified as ‘expected’ and therefore exempt from expedited reporting, although all will be recorded on a log (see Table 3). See Appendix 2 for a full list of AE descriptions and details.

Table 3 Adverse events of special interest (AESIs) and expected serious adverse events (ESAEs)
Full size table

Discussion

Despite the interest in, and potential impact of, conservative treatment options for CSDH patients, there currently exists no level 1 evidence to support any drug treatments. However, several studies have supported the use of dexamethasone and shown some evidence of its efficacy in reducing recurrence or as a primary treatment for CSDH [3,4,5, 9]. As a result, some clinicians are beginning to adopt dexamethasone as a treatment option in their routine practice. Assimilation of such new therapies into clinical care should be avoided until definitive evidence is available. The reasoning for this is exemplified by Prud’homme et al., who highlighted the potential adverse side effects associated with dexamethasone therapy in this patient population [23]. Such findings must be considered in trial design and it is evident that proving whether a medication is effective is not sufficient. One must also review the risk–benefit profile of a treatment to ensure the overall outcome affords significant benefit to the patient. Therefore, we have focused on functional outcome measures (e.g. mRS), so that the overall effect on quality of life, rather than change in imaging or tissue biomarker, is used to gauge success. This can only be achieved with a pragmatic, multi-centre trial, such as Dex-CSDH. Understanding the cost implications of new therapies is also important in NHS practice; therefore, health-economic analysis has also been incorporated in this trial protocol.

Potential limitations to this study include: reaching adequate recruitment in this patient population, as it has been highlighted that researchers can be reluctant to recruit elderly patients to randomised controlled trials [24]; and ensuring a range of severity of CSDH patients are included, as the more severely unwell patients will be unable to consent for themselves and NOK consent may be perceived as a barrier to recruitment if they are not immediately available. To help overcome this, IHP consent is an option and we hope this will be utilised to ensure broad inclusion of appropriate patients. Finally, as this study is limited to the UK, it may be questioned how applicable it is to other populations; however, we have sought to recruit centres covering a diverse range of patient demographics and therefore envisage the results will still be widely applicable.

The trial has successfully completed its feasibility phase of the first 100 patients and is now into the final phase of recruitment in 22 neurosurgical centres throughout the UK. Regular review of unblinded safety data is performed by the IDMEC, who have reported no concerns thus far. There is weekly oversight of the trial by a trial management group and biannual TSC meetings. Neurosurgical trainees have also been essential to the ongoing success of the trial through the British Neurosurgical Trainees Research Collaborative (BNTRC). This is a group that was founded in 2012 with the aim of encouraging high-quality multi-centre research within UK neurosurgery [25]. It promotes the structure where there is a trainee co-principle investigator (Co-PI) at each centre, helping oversee local site management and recruitment alongside the PI and research nurse team. The Dex-CSDH trial offers a model of how multi-centre trials can be successful in the UK with support from the wider neurosurgical community, including trainee collaboration.

Trial status

Recruitment commenced on 13 August 2015 and is ongoing under protocol version 3 (27 Apr 2017) with 630 patients recruited as of 25 June 2018 across 22 UK sites. The protocol was written in line with the SPIRIT guidelines (see Additional file 1).