Corticosteroids for aneurysmal subarachnoid haemorrhage and primary intracerebral haemorrhage
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
-
To determine whether corticosteroid therapy reduces the proportion of patients who die or have a poor outcome at one and six months after the onset of SAH or PICH.
-
To determine whether corticosteroid therapy reduces the frequency of DCI in patients with SAH.
-
To determine the frequency of adverse effects of corticosteroid therapy in patients with SAH and PICH within six months of the onset of the event.
Background
Subarachnoid haemorrhage (SAH) and primary intracerebral haemorrhage (PICH) currently account for approximately 20% of all strokes and lead to early (one month) death in about 32% and 42% patients, respectively (Bamford 1990, Broderick 1993, Feigin 2003). Delayed cerebral ischaemia (DCI) in patients with aneurysmal SAH (Ljunggren 1983; Longstreth 1993) and secondary perifocal brain oedema commonly followed by herniation from the space‐occupying effect of the PICH (Clasen 1980) remain major causes of death and poor functional outcome in these patients. It has also been suggested that advanced patient's age (>60 years), a deep location and size of the haematoma of more than 40 ml are related to poor outcome in patients with PICH (Beldman 1996), while a large amount of subarachnoid blood on CT‐scan (Brouwers 1993) and a depressed level of consciousness on admission (Hijdra 1988; Hop 1999) are important predictors for the development of DCI in patients with aneurysmal SAH.
There is evidence that decreasing plasma volume (Wijdicks 1985), hyponatraemia (Wijdicks 1985; Wijdicks 1988a), impaired autoregulation of cerebral blood flow (Ischii 1979), and reactive inflammation (Chyatte 1987) are important contributing factors to the development of DCI after aneurysmal SAH. Moreover, mineralocorticoid treatment with fludrocortisone acetate prevents plasma volume depletion (Wijdicks 1988b) and glucocorticoid treatment (methylprednisolone with or without cortisol) has an anti‐inflammatory effect (Chyatte 1983) and results in cerebral vasodilation (Fox 1975; Hashi 1980) and improvement of cerebral blood flow (Yamakawa 1991) following SAH. However, a beneficial effect of steroids on clinical outcome in patients with SAH has not been proven (Meyer 1995). Although corticosteroids may also have a beneficial effect in patients with PICH, it has been suggested that in patients with putamino‐capsular haemorrhages steroids have a harmful effect on the clinical outcome (De Reuck 1989). Moreover, many studies found an increased frequency of complications such as gastrointestinal bleeding, infection, and exacerbation of diabetes mellitus in patients treated with corticosteroids after a stroke (Bauer 1973; Tellez 1973; Norris 1976; Poungvarin 1987). Despite lack of evidence, corticosteroids, particularly dexamethasone, are commonly used for treatments in patients with SAH and PICH (Desai 1998; Karnik 1990; Kase 1991; Longstreth 1993).
In order to assess the efficacy and safety of corticosteroids in the treatment of patients with acute SAH and PICH we perform a systematic overview of all randomized controlled trials (RCTs) in this field.
Objectives
-
To determine whether corticosteroid therapy reduces the proportion of patients who die or have a poor outcome at one and six months after the onset of SAH or PICH.
-
To determine whether corticosteroid therapy reduces the frequency of DCI in patients with SAH.
-
To determine the frequency of adverse effects of corticosteroid therapy in patients with SAH and PICH within six months of the onset of the event.
Methods
Criteria for considering studies for this review
Types of studies
We seek to identify all truly and quasi‐randomised clinical trials of corticosteroid therapy in patients with SAH or PICH that have a placebo or standard strategy arm as control. Uncontrolled studies will be excluded from the analysis. Controlled trials where allocation to treatment or control group was not adequately concealed (e.g., allocation by alternation, open random number list, date of birth, day of the week or hospital number) will be included in the analysis, but analysed separately as well.
Types of participants
Patients of any age and either gender with PICH documented by CT‐scan or MRI. Patients of any age and either gender with SAH documented by CT‐scan, MRI or cerebrospinal fluid examination. These patients could be of any clinical condition before the start of corticosteroid therapy.
Types of interventions
Corticosteroid therapy of any type (mineralocorticoids, glucocorticoids) regardless of dosage and mode of administration (intravenously, orally, or both). This treatment can be started before or after aneurysm/haematoma surgery.
Types of outcome measures
The main outcomes of interest to be extracted from each treatment group are the following:
Primary outcome measures:
-
Death within the first month and after one to six months of the event onset;
-
Poor outcome (death, severe disability, or vegetative state) within the first month and after one to six months of the event onset
Secondary outcome measures:
-
Development of DCI (including clinical signs of ischaemic neurological deficit) during the scheduled treatment or follow‐up period;
-
Adverse effects of the treatment (such as hypokalemia, hyperglycemia, any infection, gastrointestinal bleeding, exacerbation of diabetes mellitus) during the scheduled treatment or follow‐up period.
Given that mineralocorticoids and glucocorticoids have different mechanisms of action (glucocorticoids have predominant action on intermediary metabolism, whereas mineralocorticoids have predominant action on the metabolism of sodium and potassium), the effect of these two groups of drugs on clinical outcome in patients with SAH and PICH will be considered for each group of drugs separately and combined in order to minimise clinical heterogeneity. We will include in the analysis the number of patients with DCI as given in the original report, regardless whether or not a definition was given in that report or whether this was adequate. Duration of the follow‐up period to be used for data extraction will be restricted to six months.
Search methods for identification of studies
This review will draw on the Search Strategy developed for the Stroke Group as a whole. Relevant trials will be identified in the Group's Specialised Register of Controlled Trials (see Review Group for more information). In addition, we will search the Cochrane Central Register of Controlled Trials (Cochrane Library), MEDLINE (1966 to present) and EMBASE (1980 to present), and search reference lists of relevant studies identified. We also plan to make an attempt to identify any relevant ongoing and unpublished studies. The following search strategy will be used for MEDLINE and modified for the other databases.
1. exp intracranial hemorrhages/
2. vasospasm, intracranial/
3. intracranial aneurysm/
4. aneurysm/or aneurysm, ruptured/ or hematoma/
5. exp brain/ or exp meninges/
6. 4 and 5
7. ((brain or intracranial or cerebral or basal ganglia or subarachnoid or intracerebral or putaminal) adj5 (haemorrhage$ or hemorrhage$)).tw.
8. ((brain or cerebral or intracerebral or intracranial or basilar or communicating artery) adj5 aneurysm$).tw.
9. ((brain or intracerebral or intracranial or subarachnoid) adj5 (haematoma$ or hematoma$).tw.
10. (SAH or PICH).tw.
11. 1 or 2 or 3 or 6 or 7 or 8 or 9 or 10
12. exp adrenal cortex hormones/
13. exp glucocorticoids, synthetic/
14. exp mineralocorticoids, synthetic/
15. (corticosteroid$ or corticoid$ or glucocorticoid$).tw.
16. (betamethasone or dexamethasone or methylprednisolone or prednisolone or triamcinolone or fludrocortisone).tw.
17. (corticosterone or hydrocortisone or hydrocorticosteroids or tetrahydrocort$ or desoxycorticosterone).tw.
18. 12 or 13 or 14 or 15 or 16 or 17
19. 11 and 18
20. limit 19 to human
Data collection and analysis
Data extraction and trial quality assessment
Three of us (VLF, DAB, GJER) will independently extract details of randomization methods, blinding of treatments and assessments, whether treatment groups were comparable with regard to major prognostic risk factors for outcomes, the number of patients who were excluded or lost to follow‐up, definition of outcomes, and entry and exclusion criteria. The methodological quality of each trial will be assessed by three raters (VLF, DAB and GJER) independently from each other as described elsewhere (Jadad 1996). The number of patients with each outcome of interest and side effect specified in the section on Types of outcome measures will also be recorded. All data will be independently extracted by the same three reviewers and cross‐checked. Any inconsistency in the extracted data between the three reviewers will be assessed and resolved by discussion (if no consensus can be reached between the three reviewers, a fourth reviewer [NA] will be approached and his decision will be final). If there will be patients excluded or lost to follow‐up after randomization or if any of the above data would not be available from the publications, further information will be sought by contacting the trialists. If the data about patients excluded or lost to follow‐up remains unavailable, a decision whether to include this particular trial in the review will be made by all authors.
Data analysis
A weighted estimate of the typical treatment effect across trials (relative risk) with absolute risk reduction by means of a standard fixed‐effects method (APT 1994) and the test for heterogeneity between trial results with a standard chi‐squared test will be calculated by means of the Cochrane statistical software, RevMan.
Sensitivity analysis
To provide an intention‐to‐treat analysis, we aim to extract from each trial the number of patients who were originally allocated to each treatment group. Any significantly positive trial which included patients excluded or lost to follow‐up (protocol analysis) will be re‐evaluated by a worst case scenario meta‐analysis. This analysis assumes that those patients who have been excluded or lost to follow‐up in the treatment group have had the worst outcome while those patients who have been excluded or lost to follow‐up in the control group have had the best outcome. Additional sensitivity analyses will be carried out with estimation of treatment effects (relative risk) by means of a random‐effect model, with exclusion of a mega‐trial, if there will be any, that contributed most to the data, and with exclusion of trials with inferior methodological quality. The presence of publication bias will be assessed by the funnel plot method (Egger 1997).
Additional subgroup or sensitivity analyses will be based on: (1) the comparison of the efficacy of steroid therapy in truly randomized trials versus all trials; (2) the exclusion of trials with standard strategy as control; (3) the comparison of the efficacy of various dosages and routes of administration of steroids; (4) the comparison of the efficacy of steroid treatment started within the first 24 hours and later than 24 hours of PICH/SAH onset. Additional analyses will be based on: (1) the comparison of the efficacy of glucocorticoids and mineralocorticoids; and (2) the comparison of the efficacy of treatment with steroids started before and after aneurysm/haematoma surgery. All these analyses will be carried out by means of the Cochrane statistical software (RevMan), as for the main data analyses.
