CT Clock Tool Feasibility Study

Background

Stroke is a major health concern worldwide causing high rates of disability and death. Most strokes (approximately 85%) are cause by ischaemia following blockage to an arterial blood vessel. Two effective treatments are available that can remove the vessel blockage and thus reduce the rates of both disability and death after stroke. However, both treatments have time-limits for safe delivery. Thrombolysis uses an injected medicine to break down the blood clot and must be used within 4.5 hours of stroke onset (the European licensing time limit for alteplase). Thrombectomy is a surgical procedure that involves pulling the clot out using a catheter placed inside the artery and is usually limited to 6 hours from stroke onset.

Unfortunately, for approximately 20% of patients with ischaemic stroke, the time of symptom onset is not known, usually because patients wake with symptoms or are found collapsed. Such patients have traditionally been denied treatment, but recent research has shown that advanced brain imaging methods (for example CT perfusion or MRI) can be used to identify patients who can still safely be treated in the absence of a known onset time for stroke. However, these advanced imaging methods are not widely available for acute stroke assessment, especially outside major treatment centres in developed nations. For example, it remains difficult in Scotland to offer advanced brain imaging to all who might need it. Plain (non-enhanced) CT is much more widely available and is the standard immediate assessment method for most patients with stroke worldwide.

As ischaemic stroke progresses, injured brain slowly begins to change on plain CT and the ischaemic lesion becomes more and more pronounced with greater oedema and swelling. These changes are seen as a darkening of tissue, indicating a decrease in CT attenuation.

The investigators have developed a method for estimating time of ischaemic stroke onset for individual patients that uses only attenuation measurements of plain CT brain imaging, the CT Clock Tool.

Rationale

The proposed CT Clock Tool method for estimating the time of stroke onset (and thus clarifying individual patient eligibility for effective treatment) is significantly simpler to use in the acute setting than the advanced methods currently available (CT perfusion or MRI).

The CT Clock Tool can provide specific time estimates for a given patient. These estimates are derived from simple measurements of brain that can be manually acquired from non-enhanced CT scans within seconds, no other imaging or specialist viewing software is required. Specifically, users (for example stroke clinicians or radiologists) measure the attenuation of any visible ischaemic lesion and of the equivalent normal brain to derive a CT attenuation ratio (attenuation of ischaemic brain ÷ attenuation of normal brain).

Initial testing shows that an attenuation ratio greater than 0.82 is highly predictive of a stroke onset to CT time less than 4.5 hours (i.e. within the European time limit for thrombolysis with alteplase). At this threshold, the CT Clock Tool is 97% sensitive and 83% specific for correctly classifying patients who are within 4.5 hours of stroke symptom onset. In addition, the investigators have tested associations between the appearances of non-enhanced CT and CT perfusion (CTP) when both are acquired at the same time. The CT attenuation ratio was compared with CTP appearances that indicate affected brain tissue is likely to remain viable. In other words, and in the absence of CTP, a CT attenuation ratio greater than 0.87 is 86% sensitive and 91% specific for identifying brain tissue that remains viable (and is therefore most likely to benefit from treatment).

The longer-term aim is to incorporate the CT Clock Tool into routine clinical care, especially in centres that cannot provide alternative or advanced methods to assess patient eligibility for acute treatment.

This current study seeks to provide a first prospective clinical feasibility evaluation to assess the practicalities of using the method in real time and to evaluate the potential impact on clinical care. The results of the present analysis will inform the design of a future randomised-controlled trial to evaluate whether the CT Clock Tool can safely be used to determine the eligibility of patients for treatment with thrombolysis and/or thrombectomy where time of stroke onset is unknown or is delayed.