Sonothrombolysis for acute ischaemic stroke
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
-
To quantify the potential benefits of sonothrombolysis in the treatment of acute ischaemic stroke.
-
To quantify the potential harms of sonothrombolysis in terms of cerebral bleeding (both haemorrhagic transformation and cerebral haemorrhage).
Background
Description of the condition
Stroke is the second most common cause of death and the major cause of disability worldwide. The proportion of deaths caused by stroke is 10% to 12% in western countries, and 12% of these deaths are in people less than 65 years of age. Advances have occurred in the prevention and treatment of stroke during the past decade. In acute settings, intravenous recombinant tissue plasminogen activator (r‐tPA) has been shown to be effective in selected patients with acute ischaemic stroke within three hours from symptom onset, and its use is spreading in stroke centres (Donnan 2008; Wardlaw 2003).
Moreover, the ECASS III study results confirmed that intravenous r‐tPA in acute ischaemic stroke between three and 4.5 hours after the onset of symptoms significantly improved clinical outcomes in patients with acute ischaemic stroke (Hacke 2008). Finally, the IST‐3 study is currently evaluating the possibility of extending the therapeutic window in terms of time, age and severity (Sandercock 2008).
Description of the intervention
Several studies have consistently demonstrated the capability of ultrasound to enhance the lysis of intra‐arterial thrombus in acute ischaemic stroke during systemic intravenous thrombolysis with tPA, an intervention also called sonothrombolysis (Braaten 1997; Ishibashi 2002; Kimura 1994). Ultrasound has also been used during intra‐arterial thrombolysis but we will not deal with this technique in our review (Mahon 2003; Tomsick 2008).
Sonothrombolysis is usually performed in patients with documented arterial occlusion by insonation of major intracranial vessels, including the middle cerebral artery and intracranial internal carotid artery. Ultrasound‐enhanced thrombolysis can be further amplified by adding gaseous microbubbles of ultrasound contrast agents used in neurovascular imaging (Molina 2006).
Even in the absence of thrombolytic drugs, ultrasound has been demonstrated to induce clot lysis, which is probably related to the improvement of intrinsic enzymatic fibrinolysis (Cintas 2002).
The length of application of ultrasound therapy is likely to be variable, in that it can last from less than 30 minutes to two hours.
Middle cerebral artery and internal carotid artery occlusions can be treated with ultrasound delivered through temporal bone windows towards the acute arterial occlusion in different ways, which include:
-
diagnostic transcranial doppler (TCD) at 2 MHz frequency;
-
B‐mode/colour flow duplex imaging (TCCD) at 2 to 4 MHz frequency;
-
low‐frequency ultrasound at frequency of 300 kHz.
Moreover, contrast agents can be administered by bolus infusions or by intravenous continuous infusion.
How the intervention might work
Ultrasound increases the transport of tPA into the thrombus, promotes the opening and cleaving of the fibrin polymers, and improves the binding affinity of tPA to fibrin (Ishibashi 2002). It accelerates enzymatic fibrinolysis primarily through non‐thermal mechanisms by increasing the transport of drug molecules into the clot. Mechanical effects of ultrasound radiation forces have the ability to influence drug transport. In addition, ultrasound can promote the motion of fluid through and around the thrombus, an effect called streaming.
Microbubbles are small, gas‐filled microspheres with specific acoustic properties that make them useful as ultrasound contrast agents in neurovascular imaging. Application of high‐acoustic‐pressure ultrasound has been shown to induce non‐linear oscillations of microbubbles, leading to a continuous absorption of energy until the bubbles explode, releasing the absorbed energy. Thus, ultrasound‐mediated microbubble destruction may further accelerate the clot‐dissolving effect of ultrasound (Culp 2004; Dijkmans 2004; Tachibana 1995). Microbubbles have been shown to damage the clot surface directly, promoting microfragmentation and high‐power jetting into the clot. This mechanism is known as inertial cavitation. Moreover, it has been demonstrated that microbubbles have their own ability to lyse thrombi, even in the absence of a lytic drug (Unger 2002).
On the other hand, microbubble activation has also been shown to create endothelial damage, to promote blood barrier disruption and microvascular bleeding in animal models (Meairs 2007).
Contrast ultrasound with air‐filled albumin microbubbles has been shown experimentally to produce alterations of endothelial permeability, haemolysis and bleeding, in proportion to the intensity and duration of ultrasound exposure and to the dose of the contrast agent. The leading hypothesis is that this is probably caused by the enhanced cavitational effects.
The shock waves generated by the bubbles' explosion can induce endothelial injury: this damage mechanism is thought to be created by the collision between expanded microbubbles and the vascular wall, causing microvessel rupture and subsequent haemorrhage (Miller 1998).
Why it is important to do this review
Sonothrombolysis is a promising new feasible therapeutic intervention in acute stroke, but it is still under investigation: the ongoing debate is focused on its real efficacy and safety in enhancing the effect of tPA, technical aspects of ultrasound administration, the association with microbubbles, and the clinical inclusion criteria of the patients to be treated.
TCD and TCCD use different ultrasound frequencies (2 MHz versus 2 to 4 MHz). There are no reports in the literature about possible different biological effects of the two frequencies. As both TCD and TCCD are routinely used for diagnostic purposes, the demonstration of the efficacy of sonothrombolysis could be easily transferred to a low‐cost, bedside therapeutic application.
Considering that TCD can be continuously applied by means of a head frame, while TCCD must be hand‐held, there may be a difference in clinical handling.
Enhancing the effect of ultrasound with microbubbles could be an interesting advance in sonothrombolytic treatment, but there are still some concerns, specifically about the best method of infusion (bolus or continuous infusion) and the risk of haemorrhagic transformation.
Furthermore, when microbubbles are administered by repetitive boli for diagnostic purposes during ultrasound application for sonothrombolysis, the administration could be considered as sonothrombolysis enhanced by microbubbles. There are some reports of endothelial damage and blood‐brain barrier disruption in patients treated with sonothrombolysis associated with microbubbles.
However, data regarding the effect of ultrasound‐activated microbubbles on an already disrupted blood‐brain barrier in the course of cerebral ischaemia are lacking. Recently, experimental studies showed that ultrasound neither promoted additional blood‐brain barrier disruption nor increased apoptosis and markers of tissue damage outside the infarcted area (Stroick 2006).
These conflicting experimental data underline the possible harm of both ultrasound and microbubbles in acute ischaemic stroke.
Objectives
-
To quantify the potential benefits of sonothrombolysis in the treatment of acute ischaemic stroke.
-
To quantify the potential harms of sonothrombolysis in terms of cerebral bleeding (both haemorrhagic transformation and cerebral haemorrhage).
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials with clear allocation concealment.
Types of participants
Patients admitted to hospital, with definite acute ischaemic stroke (diagnosis made by computerised tomography (CT) scan or magnetic resonance (MR), with treatment started within 12 hours from symptom onset).
Types of interventions
Sonothrombolysis (any duration and any frequency, with or without microbubbles) versus intravenous tPA therapy alone or conventional treatment.
Comparisons
Our main comparison will be all patients treated with sonothrombolysis versus all patients not treated with sonothrombolysis. We will also look at the following comparisons:
-
all patients treated with sonothrombolysis versus those not treated with sonothrombolysis, tPA given to both groups;
-
all patients treated with sonothrombolysis versus those not treated with sonothrombolysis, but tPA not given;
-
all patients treated with sonothrombolysis and microbubbles versus all patients treated with sonothrombolysis but without microbubbles;
-
all patients treated with sonothrombolysis and microbubbles versus all patients treated with sonothrombolysis but without microbubbles in patients treated with tPA;
-
all patients treated with sonothrombolysis and microbubbles versus all patients treated with sonothrombolysis but without microbubbles in patients not treated with tPA.
Types of outcome measures
Primary outcomes
Survival free of significant disability, measured with the Oxford Handicap Scale (OHS) or equivalent, at three to six months from stroke. When the OHS is available, values greater than two will be considered as showing significant disability; when equivalent scales are available, we will classify as 'significant disability' a result which is as similar as possible to OHS greater than two.
Secondary outcomes
Mortality, recanalisation (determined by any method), symptomatic and asymptomatic haemorrhagic transformation, and cerebral haemorrhage (detection of blood in sites different from infarction location) (Berger 2001; Hacke 1995; Larrue 2001; Wahlgren 2007).
Search methods for identification of studies
See the 'Specialized register' section in the Cochrane Stroke Group module.
Electronic searches
We will search the Cochrane Stroke Group Trials Register. In addition, we will search the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, latest issue), MEDLINE (1950 to present) (Appendix 1), and EMBASE (1980 to present).
Searching other resources
In an effort to identify further published, unpublished and ongoing studies:
-
we will search the Database of Abstracts and Reviews of Effects (DARE) (The Cochrane Library, latest issue), in an attempt to identify relevant systematic reviews and meta analyses;
-
we will search the following ongoing trials and research registers: Stroke Trials Registry (http://www.strokecenter.org/trials/), Clinicaltrials.gov (http://clinicaltrials.gov/), and Current Controlled Trials (http://www.controlled‐trials.com/);
-
we will search reference lists from relevant articles and reviews;
-
we will attempt to identify and handsearch relevant journals and conference proceedings not already searched on behalf of The Cochrane Collaboration;
-
we will contact colleagues, authors and researchers active in the field, and companies involved in relevant research and development.
We will search for relevant trials in all languages and, when necessary, arrange translation of trial reports published in languages other than English or Italian.
Data collection and analysis
Selection of studies
Three review authors (LD, TM, SC) will independently select potentially eligible studies identified by the search. If, from the title or abstract, it is clear that a study is not suitable for inclusion in the review (e.g. a non‐randomised study), we will exclude it. In case of doubt, we will examine the full text. The three review authors will resolve any disagreements by discussion. We will extract study characteristics using a predefined form and include an assessment of quality. We will hold a consensus meeting of all the authors to resolve any disagreement in the quality assessment of the trials.
Data extraction and management
Two review authors (TM and LD) will independently extract data using a predefined extraction form. We will hold a consensus meeting of all the authors to deal with differences in the extracted data. If necessary, we will contact study authors for additional unpublished data.
We will extract the following data:
-
general information: published/unpublished, title, authors, country, year of publication, duplicate publication;
-
trial characteristics: design, duration, allocation concealment (and method), randomisation (and method), blinding (outcome assessors), checking of blinding, intention‐to‐treat analysis;
-
intervention: type of sonothrombolysis, duration, use of microbubbles and type, use of tPA;
-
participants: exclusion criteria, total number and number in comparison groups, age, gender, similarity of groups at baseline, severity of stroke, withdrawals/losses to follow up;
-
outcomes: listed above.
Assessment of risk of bias in included studies
We will assess the methodological quality of selected studies as described in section 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008). We will score each of the following points as 'yes', 'no', or 'unclear' (where 'yes' indicates that the study is less open to bias) and report them in the 'Risk of bias' tables.
-
Method of randomisation (selection bias): methods of randomisation using data of birth, date of admission, hospital numbers, or alternation are not appropriate because they do not give each study participant the same chance of receiving each intervention.
-
Concealment of allocation (indication bias): adequate measures to conceal allocation are central randomisation; serially numbered, opaque, sealed envelopes; or other descriptions with convincing concealment.
-
Blinding of investigators and patients (performance bias).
-
Blinding of outcome assessment (detection bias).
-
Adequate follow up (attrition bias): we will carefully check the reporting of withdrawals, drop‐outs, protocol deviations, and losses to follow up; we will consider follow up to be adequate when more than 90% of patients are reported.
-
Other possible bias.
On the basis of these criteria, we will divide studies into the following three categories.
-
A ‐ all quality criteria met: low risk of bias.
-
B ‐ one or more of the quality criteria only partly met: moderate risk of bias.
-
C ‐ one or more criteria not met: high risk of bias.
To avoid selection bias, we will not reject any study because of methodological characteristics or any subjective quality criteria, except non‐randomised studies. However, we will conduct sensitivity analyses to examine differences in study methods.
Measures of treatment effect
We will use the Cochrane Review Manager software to analyse the data (RevMan 2008). We will base the quantitative analysis of outcome on an intention‐to‐treat principle. We will use the odds ratio (OR) with 95% confidence interval (CI) to measure treatment effect for each study.
Unit of analysis issues
Due to the specificity of the treatment we are studying, we do not expect any specific problems.
Dealing with missing data
If data are missing, we will contact the original investigators for additional information. If some data remain unavailable, we will consider both best‐case and worst‐case scenarios.
Assessment of heterogeneity
We will determine heterogeneity using the I2 statistic. We will consider I2 greater than 50% to be substantial heterogeneity. If there is substantial heterogeneity, we will look for the potential sources of the heterogeneity (i.e. different types of study participants, different study designs).
Assessment of reporting biases
We will use the funnel plot method.
Data synthesis
We will estimate the overall treatment effect by the pooled OR with 95% CI using both a fixed‐effect and a random‐effects model. Each test for significance will be two‐sided.
Subgroup analysis and investigation of heterogeneity
As stated in the 'Comparisons' section above, we will examine:
-
all patients treated with sonothrombolysis versus those not treated with sonothrombolysis, tPA given to both groups;
-
all patients treated with sonothrombolysis versus those not treated with sonothrombolysis, but tPA not given;
-
all patients treated with sonothrombolysis and microbubbles versus all patients treated with sonothrombolysis but without microbubbles;
-
all patients treated with sonothrombolysis and microbubbles versus all patients treated with sonothrombolysis but without microbubbles in patients treated with tPA;
-
all patients treated with sonothrombolysis and microbubbles versus all patients treated with sonothrombolysis but without microbubbles in patients not treated with tPA.
We also plan to examine the concomitant use of aspirin.
Sensitivity analysis
We will use both fixed‐effect and random‐effects models. As already described, we will analyse differences in study methodology. We will also look at differences in intensity and duration of sonothrombolysis.
