Imaging the ischaemic penumbra with T 2 * weighted MRI

The positive results from the recent multicentre randomised clinical trials of mechanical thrombectomy have at long last reinvigorated the stroke community and open up new possibilities for protective strategies dealing with reperfusion-based injury. With efficacy demonstrated up to 6 h after stroke onset and with some patients in the ESCAPE trial ¹ being treated beyond 6 h, it raises the possibility of an extended time window in a select group of patients. In order for this to be the case, there is a need for advanced imaging techniques to identify patients with viable ischaemic penumbra if treatment (i.e. alteplase, thrombectomy and/or new protective strategies) is to be given, particularly outside the current time window.

The recent study by Shen et al. ² , provides further validation of a promising MRI technique for identifying the ischaemic penumbra based on tissue metabolism. This technique utilises inhalation of oxygen (oxygen challenge, OC) during T₂* weighted MRI to exploit differences in deoxy/oxyhaemoglobin ratio within ischaemic core, penumbra and normal tissue. Within the ischaemic penumbra, where oxygen extraction fraction and therefore deoxyhaemoglobin ratios are elevated, the OC results in a greater T₂* signal change. Originally developed by Santosh et al., ³ this technique has been evaluated in a number of pre-clinical stroke studies to confirm that tissue displaying an increased T₂* signal change to OC displays morphologically normal neurones, glucose metabolism, reduction in lactate during OC and recovery following reperfusion.^3–7 Deuchar et al. ⁸ further developed the technique using intravenous oxygen carriers to amplify the T₂* signal change and remove potential confounds of 100% oxygen in the clinic. ⁸

One of the limitations of the OC technique is that large veins and venules will display an increased T₂* signal change, similar to that observed in the ‘ischaemic penumbra’ thereby making it difficult to distinguish these non-specific changes. Shen et al. ² attempt to overcome this by investigating the time to peak (TTP) of the OC signal in addition to the amplitude of signal change. They demonstrated that tissue at risk (i.e. penumbra) can be better stratified with a delayed TTP in addition to an amplified T₂* signal during OC and that this tissue showed good correspondence with tissue at risk identified with perfusion-diffusion mismatch MRI. In contrast, large veins and venules did not show a delayed TTP. Initiation of reperfusion returned the TTP to normal and attenuated the OC T₂* signal change (within tissue at risk). Importantly, this tissue was not incorporated into the infarct when followed up at 24 h post-stroke.

This development represents a further refinement in the OC technique for identifying tissue at risk following stroke. In view of positive animal data from two independent groups, plus preliminary data in acute stroke patients, ⁹ the potential for successful translation looks promising. Further validation of the technique, such as establishing appropriate thresholds, is required before it would likely be implemented as a diagnostic technique for acute ischaemic stroke management.